Ranking

Tasks (generated 2026-10-30; regenerated by the checker, never edited)

Spell-out and agent work

• spell-out candidates (unspelled frames and childless wants): 13; heaviest normative-record, want-address-anti-naturality, want-basin-of-attraction-formalized
• agent work the map wants most: do item-130, do ddb-translation-table, do item-132, do tt-mart-repair-via-bounds-transfer, do vc-route-hedged-installation
• proposals awaiting approval: none

Lines

• commit line: research first: the decision term (printed, not applied) favours vet hull-condition-total-trust (ranked unit 43), D 0.00016
• regime: the decision term would reorder 1286 units (largest D 0.00017); shown, not applied
• act: the act's leading option is inter-inductor-total-trust → inter-inductor-characterized at value 1.000; resting on AI numbers only; derived P 1.000 against its gestalt 0.53
• audit (36 min of this session): re-perform estimate anson-002 by @anson on {"row": ["undescribed-failure", "uf-high-conflict"], "field": "p"}; cross-audit preferred (the session's audit share, listed first and never ranked; the list's top task is ranked unit 1, the commit line is about which plan to commit to)

Research and whole-argument units

The three best-ranked units of each kind that produces new structure or reads an argument whole (improve, a human's do, re-examine, spell-out, connect, revive), each numbered by its rank in the list below; the order is the list's, this section only gathers them (signals hoisted below are dropped here too).

1. spell-out normative-record (S) path: top → the-program → normativity → normative-record
2. re-examine fa-trust-half-weak-grades.improvable (S) — vet this AI number; promise rests on AI judgment path: top → the-program → corrigibility-landscape → deference-and-trust → sams-conjecture-li → fa-trust-half-weak-grades.improvable
3. re-examine uea-floored-agent-theorem (M) — vet this AI number; promise rests on AI judgment path: top → the-program → stance-and-level → level-super-turing → uea-floored-agent-theorem
4. re-examine li-compatible-abstraction-theory (M) — structure suspect: derived 0.76 against gestalt 0.42; vet this AI number path: top → the-program → abstraction-arc → stage5-li-representation-theorem → li-compatible-abstraction-theory
5. improve route-approximate-dd (S) — vet this AI number; promise rests on AI judgment path: top → the-program → decision-theory → updateless-dt-programme → udt-rep-improvable → route-approximate-dd relax an assumption, change the formalism, take an informal direction, or critique and redraft route-approximate-dd
6. improve route-interval-resolution (S) — vet this AI number; promise rests on AI judgment path: top → the-program → decision-theory → updateless-dt-programme → bli-picture-improvable → route-interval-resolution relax an assumption, change the formalism, take an informal direction, or critique and redraft route-interval-resolution
7. improve route-state-dependent-exploration (S) — vet this AI number; promise rests on AI judgment path: top → the-program → decision-theory → updateless-dt-programme → formal-core-improvable → route-state-dependent-exploration relax an assumption, change the formalism, take an informal direction, or critique and redraft route-state-dependent-exploration
8. do infinite-frames-sentence-check (S) — vet this AI number; promise rests on AI judgment path: top → the-program → stance-and-level → level-above-quantifiable → infinite-frames-sentence-check Read value_imp_totalTrust and the Coin and Bentham rows of lit-weatherson-frames-report.md, and say whether li-deference.md's sentence on infinite frames needs revising.
9. do value-uncertainty-is-computational (M) — vet this AI number; promise rests on AI judgment path: top → the-program → stance-and-level → value-uncertainty-is-computational Write the canonical-statement page from journal 2026-03-26 and the value-change note's §2.1 and §2.7, recording that the event-algebra representation is already worked and where it reaches the logical-induction agent.
10. do distributed-cooperative-oracles-route (S) — for coordination-with-others — vet this AI number; promise rests on AI judgment path: top → the-program → decision-theory → updateless-dt-programme → coordination-with-others → distributed-cooperative-oracles-route Check whether distributed cooperative oracles compose with logical induction to give updateless agents a coordination procedure, starting from the coordination phase of the 2024 agenda.
11. spell-out want-address-anti-naturality (S) — owner has not confirmed path: top → the-program → corrigibility-landscape → anti-naturality-review → want-address-anti-naturality
12. spell-out want-basin-of-attraction-formalized (S) — owner has not confirmed path: top → the-program → corrigibility-landscape → want-basin-of-attraction-formalized
13. connect want-corrigibility-theorems-via-legitimacy (S) — owner has not confirmed path: top → the-program → legitimacy → want-corrigibility-theorems-via-legitimacy
14. connect want-legitimacy-anti-meme-post (M) — owner has not confirmed path: top → the-program → legitimacy → want-legitimacy-anti-meme-post
15. connect want-methodology-note (S) — owner has not confirmed path: top → the-program → stance-and-level → level-super-turing → want-methodology-note

Ranked units (by map rate; D under --why, never applied)

The order: what each unit's result would tell about the goals it informs, in bits, each goal weighted by its stake (the plan, want or claim it serves) times its promise (how likely it looks, plus a summand where its derived number and its gestalt disagree), per hour of the unit's work (PROTOCOL §6). Each unit: n. kind target (size)[ — for <the goal carrying the largest share of its weighted bits, where that is not the target itself>][; also <goals at a fifth or more>][ — signals], then its path from the top frame (→ part of, ⇢ evidence for; the chain shown is the shortest), then the work itself where the unit names one.

1. gestalt traderized-finite-date-conformance (M) — round; work order #48 path: top → the-program → normativity → operative-force → traderized-finite-date-conformance
2. gestalt traderized-finite-date-conformance.improvable (M) path: top → the-program → normativity → operative-force → traderized-finite-date-conformance.improvable
3. gestalt verdict-licence-chain (M) — round; round path: top → the-program → normativity → normative-inductor → verdict-licence-chain
4. gestalt verdict-licence-chain.improvable (M) path: top → the-program → normativity → normative-inductor → verdict-licence-chain.improvable
5. gestalt split-bill-is-robust-openness (M) — round; round path: top → the-program → legitimacy → openness-and-non-capture → split-bill-is-robust-openness
6. gestalt split-bill-is-robust-openness.improvable (S) path: top → the-program → legitimacy → openness-and-non-capture → split-bill-is-robust-openness.improvable
7. gestalt effective-authority-layer (L) — round; round path: top → the-program → legitimacy → authorship-and-transparency → effective-authority-layer
8. gestalt effective-authority-layer.improvable (M) path: top → the-program → legitimacy → authorship-and-transparency → effective-authority-layer.improvable
9. gestalt kernel-box2-finite-time (M) — hypotheses unread; round path: top → the-program → legitimacy → formalism-open-questions-0-3 → kernel-box2-finite-time
10. gestalt local-delegation-bridge (M) — round; round path: top → the-program → corrigibility-landscape → deference-and-trust → deference-finite-results → local-delegation-bridge
11. gestalt local-delegation-bridge.improvable (M) path: top → the-program → corrigibility-landscape → deference-and-trust → deference-finite-results → local-delegation-bridge.improvable
12. gestalt effective-authority-layer.residual (S) path: top → the-program → legitimacy → authorship-and-transparency → effective-authority-layer.improvable → effective-authority-layer.residual
13. gestalt item-129 (S) path: top → the-program → normativity → operative-force → traderized-finite-date-conformance.improvable → item-129
14. gestalt item-130 (S) path: top → the-program → legitimacy → openness-and-non-capture → split-bill-is-robust-openness.improvable → item-130
15. gestalt item-131 (S) path: top → the-program → legitimacy → authorship-and-transparency → effective-authority-layer.improvable → item-131
16. gestalt item-132 (S) path: top → the-program → corrigibility-landscape → deference-and-trust → deference-finite-results → local-delegation-bridge.improvable → item-132
17. gestalt local-delegation-bridge.residual (S) path: top → the-program → corrigibility-landscape → deference-and-trust → deference-finite-results → local-delegation-bridge.improvable → local-delegation-bridge.residual
18. gestalt route-authority-attention-physics (S) — work order #101 path: top → the-program → legitimacy → authorship-and-transparency → effective-authority-layer.improvable → route-authority-attention-physics
19. gestalt route-authority-blackwell-converse (S) — work order #107 path: top → the-program → legitimacy → authorship-and-transparency → effective-authority-layer.improvable → route-authority-blackwell-converse
20. gestalt route-bridge-uniform-form (S) — work order #23 path: top → the-program → corrigibility-landscape → deference-and-trust → deference-finite-results → local-delegation-bridge.improvable → route-bridge-uniform-form
21. gestalt route-conformance-credal-constraint (S) — work order #39 path: top → the-program → normativity → operative-force → traderized-finite-date-conformance.improvable → route-conformance-credal-constraint
22. gestalt route-conformance-efficient-rows (S) — work order #42 path: top → the-program → normativity → operative-force → traderized-finite-date-conformance.improvable → route-conformance-efficient-rows
23. gestalt route-conformance-exact-and-safe (S) — work order #43 path: top → the-program → normativity → operative-force → traderized-finite-date-conformance.improvable → route-conformance-exact-and-safe
24. gestalt route-licence-authorization-from-account (S) — work order #56 path: top → the-program → normativity → normative-inductor → verdict-licence-chain.improvable → route-licence-authorization-from-account
25. gestalt route-licence-transport (S) — work order #115 path: top → the-program → normativity → normative-inductor → verdict-licence-chain.improvable → route-licence-transport
26. gestalt split-bill-is-robust-openness.residual (S) path: top → the-program → legitimacy → openness-and-non-capture → split-bill-is-robust-openness.improvable → split-bill-is-robust-openness.residual
27. gestalt traderized-finite-date-conformance.residual (S) path: top → the-program → normativity → operative-force → traderized-finite-date-conformance.improvable → traderized-finite-date-conformance.residual
28. gestalt verdict-licence-chain.residual (S) path: top → the-program → normativity → normative-inductor → verdict-licence-chain.improvable → verdict-licence-chain.residual
29. vet crux-illegitimate-ordering (S) — also for kernel-lexical-order-hypothesis — vet this AI number; promise rests on AI judgment path: top → the-program → corrigibility-landscape → legitimacy-weighted-deference → crux-illegitimate-ordering
30. vet uea-floored-agent-theorem.bridge (S) — for uea-floored-agent-theorem — attribution unvetted; vet this AI number path: top → the-program → stance-and-level → level-super-turing → uea-floored-agent-theorem → uea-floored-agent-theorem.bridge
31. vet uea-floored-agent-theorem.complete (S) — for uea-floored-agent-theorem — vet this AI number; promise rests on AI judgment path: top → the-program → stance-and-level → level-super-turing → uea-floored-agent-theorem → uea-floored-agent-theorem.complete
32. vet tt-mart-repair-via-bounds-transfer (S) — vet this AI number; promise rests on AI judgment path: top → the-program → legitimacy → deference-toward-bounded-agents → tt-mart-repair-via-bounds-transfer
33. vet vingean-uncertainty-is-computational (S) — attribution unvetted; vet this AI number path: top → the-program → stance-and-level → level-super-turing → vingean-uncertainty-is-computational
34. vet reachable-corrective-control-negative (S) — vet this AI number; promise rests on AI judgment path: top → the-program → corrigibility-landscape → public-corrigibility-line → reachable-corrective-control-negative
35. vet summable-implies-average-with-margin (S) — vet this AI number; promise rests on AI judgment path: top → the-program → corrigibility-landscape → deference-and-trust → ddb-to-li-translation → ladder-chain-as-written.improvable → route-ladder-average-form ⇢ summable-implies-average-with-margin
36. vet deference-protects-from-outprediction-not-deception (S) — vet this AI number; promise rests on AI judgment path: top → the-program → abstraction-arc → stage6-dt-and-scale-free-agency → cooperative-equilibria-as-agents → deference-protects-from-outprediction-not-deception
37. vet route-v1-timely-fragment-certificate (S) — vet this AI number; promise rests on AI judgment path: top ⇢ route-v1-timely-fragment-certificate
38. vet ladder-chain-bounded-gate (S) — vet this AI number; promise rests on AI judgment path: top → the-program → corrigibility-landscape → deference-and-trust → ddb-to-li-translation → ladder-chain-as-written.improvable → route-ladder-bounded-gate ⇢ ladder-chain-bounded-gate
39. vet published-statement-holds (S) — vet this AI number; promise rests on AI judgment path: top → the-program → corrigibility-landscape → deference-and-trust → published-statement-holds
40. vet legitimacy-blocks-wireheading (S) — vet this AI number; promise rests on AI judgment path: top → the-program → corrigibility-landscape → legitimacy-detection → legitimacy-blocks-wireheading
41. … 1812 more ranked units in derived.json (--all prints them)

Agent work

Work an agent can do without a human; the map's ordering, not PRIORITIES.md's.

1. do item-130 (S) — also for split-bill-is-robust-openness.improvable path: top → the-program → legitimacy → openness-and-non-capture → split-bill-is-robust-openness.improvable → item-130 State an intervention semantics for Robust Openness indexed by the accounted state (the docket, the ports and the accounts), producing for each state and concern the actual coverage data and the data under each declared intervention, and check the non-capture certificate against it.
2. do ddb-translation-table (S) — for ddb-reading-translation-loop — vet this AI number; promise rests on AI judgment path: top → the-program → corrigibility-landscape → deference-and-trust → ddb-to-li-translation → ddb-reading-translation-loop → ddb-translation-table Write wiki/DDB-Translation-Table.md, one row per DDB result, from the wiki pages and the clean room's lit-ddb-frames, lit-ddb-facts and lit-ddb-accuracy-mm packages, and link it from wiki/Home.md.
3. do item-132 (S) path: top → the-program → corrigibility-landscape → deference-and-trust → deference-finite-results → local-delegation-bridge.improvable → item-132 Find a settlement instantiation of the finite kernel that produces grade trust at some level η from the trajectory itself, or prove that none in the skeleton does, stating which quantity the grade would have to track.
4. do tt-mart-repair-via-bounds-transfer (S) — vet this AI number; promise rests on AI judgment path: top → the-program → legitimacy → deference-toward-bounded-agents → tt-mart-repair-via-bounds-transfer Rewrite the pending-rewrite note of projects/deference/note-dump-2026-08-11/wiki/total-trust-implies-mart.md into the true-setting proof by bounds transfer, citing tower_instance_of_totalTrust_gapBets.
5. do vc-route-hedged-installation (S) — vet this AI number; promise rests on AI judgment path: top → the-program → corrigibility-landscape → value-change-base-layer → value-change-is-belief-change.improvable → vc-route-hedged-installation Build a two-world fixture in which the correction process installs the mixture state rather than the legitimacy-conditional state, and check whether the base-rate condition disappears as the note's section 2.6 predicts.
6. do item-131 (S) path: top → the-program → legitimacy → authorship-and-transparency → effective-authority-layer.improvable → item-131 Wire the ask-list deviation into the record's Counted as a dimension of the trace, as Consult2 wired the raise dimension, so that asking off the protocol is counted below D − ϖ by the record and not only flagged as a self-checkable clause beside it.
7. do item-129 (S) path: top → the-program → normativity → operative-force → traderized-finite-date-conformance.improvable → item-129 Test whether the normativity line's constraint compiler can enforce a per-day gap-bet bound (the novice's expectation of each signed gap bet at least minus a schedule) while preserving the criterion, and state the obstruction if not.
8. do biconvex-conditioning-in-li (S) — for per-sentence-vs-whole-belief — vet this AI number; promise rests on AI judgment path: top → the-program → corrigibility-landscape → deference-and-trust → ddb-to-li-translation → per-sentence-vs-whole-belief → biconvex-conditioning-in-li Check whether 'the expert's quotes lie in B' is a market-generable weight for a biconvex set B, and record the short argument either way.
9. do distinct-novice-paper-expert (M) — vet this AI number; promise rests on AI judgment path: top → the-program → corrigibility-landscape → deference-and-trust → li-deference-collapse.improvable → route-collapse-distinct-expert ⇢ distinct-novice-paper-expert Finish def-squeeze-diamond's distinct-novice transport: the conditional-tower and Value corners over a distinct novice process and the Regularity targets 6b–6d, per its handoff.
10. do frames-beyond-finite (M) — vet this AI number; promise rests on AI judgment path: top → the-program → stance-and-level → level-above-quantifiable → ddb-value-iff-total-trust.improvable → frames-beyond-finite Record what lit-weatherson-frames proves against what Weatherson's paper claims, and list the countable-frame conditions under which Total Trust implies Value is restored.
11. do frames-with-self-reference (M) — vet this AI number; promise rests on AI judgment path: top → the-program → stance-and-level → level-above-quantifiable → ddb-value-iff-total-trust.improvable → frames-with-self-reference Write the two-page comparison of the random-variable representation against the Diagonal Lemma: what the logical-induction translation already does about self-reference, and what Christiano's framework would add.
12. do self-aixi-eps-exploration-route (M) — vet this AI number; promise rests on AI judgment path: top → the-program → stance-and-level → level-super-turing → uea-theorem-1.improvable → self-aixi-eps-exploration-route Read Kim and Lee and write a one-page comparison of the hypotheses of their exploration route, Wyeth's oracle route and the floor of Theorem C.
13. do item-110 (M) — vet this AI number; promise rests on AI judgment path: top → the-program → legitimacy → authorship-and-transparency → public-legitimacy-improvable → item-110 Execute PRIORITIES item 110: define the baseline (candidate: the idle-agent rollout read as a process on her record), re-derive the eight-world model's seven cases from it, and include the sealed sibling of the endpoint-preservation model as a second candidate for comparison.
14. do uea-current-node-floor-repair (M) — vet this AI number; promise rests on AI judgment path: top → the-program → stance-and-level → level-super-turing → uea-theorem-1.improvable → uea-current-node-floor-repair Restate Theorem 1 for the current-node-floored agent as a theorem over every reflective oracle in rOSI, with the finite Theorem C as its shadow.
15. do hull-condition-total-trust (M) — vet this AI number; promise rests on AI judgment path: top → the-program → corrigibility-landscape → legitimacy-detection → hull-condition-total-trust Prove finite-dimensional Minkowski-Weyl (or finitely generated Farkas) in Lean once and close Hull.hullCondition_of_defers. (open statement: Hull.hullCondition_of_defers)
16. do ss-value-without-scope (M) — vet this AI number; promise rests on AI judgment path: top → the-program → corrigibility-landscape → deference-and-trust → li-deference-collapse.improvable → route-collapse-scheduled-grade ⇢ ss-value-without-scope Write out the F1-for-lookahead near-identity and stress-test scheduled weighted-average Value against total-trust-implies-value's selection-punishing menu.
17. do uea-rosi-lift (M) — vet this AI number; promise rests on AI judgment path: top → the-program → stance-and-level → level-super-turing → uea-floored-agent-theorem.improvable → uea-rosi-lift Type the rOSI statements of Theorem 1 and of Theorems B and C with discounting explicit, and mark which finite-shadow steps use finiteness.
18. do value-symmetric-comparator (M) — vet this AI number; promise rests on AI judgment path: top → the-program → stance-and-level → level-above-quantifiable → ddb-value-iff-total-trust.improvable → value-symmetric-comparator Define symmetric Value on finite frames and test DDB's Fact 2.1 frame and the anti-expert frame against it. (work order: PRIORITIES #27)
19. do uea-updateless-variant (M) — vet this AI number; promise rests on AI judgment path: top → the-program → stance-and-level → level-super-turing → uea-theorem-1.improvable → uea-updateless-variant Reconcile notes/01-interleaving-sketch.md with sink-or-swim findings F-6 to F-9 and state which updateless optimality notion the repaired theorem of the updateless self-game uses.
20. do value-on-agency (M) — vet this AI number; promise rests on AI judgment path: top → the-program → corrigibility-landscape → legitimacy-weighted-deference → value-on-agency Write a first formal sketch: a value variable for human control over a decision class, learned early, compared with the kernel's lexical authority term on Box 1's house-sale instance.

By goal

The goals in the frame tree's order (the agenda's outline): kind and role, the best ranked unit and its rank, the count of units, and the Statement's first clause. A rank past the printed list is in derived.json (each unit's rank), and --all prints every ranked unit.

• the-program (frame, owner @anson, @abram): How can a powerful AI develop in line with humanity's interests without relying on control — and how can we get mathematical guarantees that it does?
  • normativity (frame, owner @anson; holds 36; owner has not confirmed): What does learning from reasons require of a reasoner: which statics a normative seed forces, how a constraint acquires operative force in a market, and what the Normative Inductor, from seed through reasons, compile, enforcement and revision, establishes end to end?
  • seed-theory (frame, owner @anson; holds 16; owner has not confirmed): What does a normative seed force on the prices of a settled fragment, how does that change under settlement, defeat and rebuttal, and do reasoners starting from different seeds end up anywhere together?
  • operative-force (frame, owner @anson; holds 27; owner has not confirmed): How does a constraint acquire operative force in a logical inductor's market without destroying the property that nobody can pump the reasoner, and what does that force cost?
  • traderized-finite-date-conformance (shared claim of operative-force) — best unit: 1330. vet compiler.deductive-end-to-end.bridge; 16 unit(s); A constraint on prices compiles to a legal trading strategy inside the market recursion
  • traderized-finite-date-conformance.improvable (improvement claim on traderized-finite-date-conformance, shared claim of operative-force) — best unit: 827. improve route-conformance-credal-constraint; 5 unit(s); The compiled per-date conformance, in which a constraint on prices compiles to a legal trading strategy that conforms within a schedule while preserving the criterion, can be improved until it…
  • item-129 (route under traderized-finite-date-conformance.improvable) — best unit: none ranked; 1 unit(s); The compiled enforcer's per-day gap bets are bounded below by a schedule while the criterion is preserved
  • route-conformance-credal-constraint (route under traderized-finite-date-conformance.improvable) — best unit: 827. improve route-conformance-credal-constraint; 1 unit(s); Normative statics produce a credal constraint on prices rather than only a price demand
  • route-conformance-efficient-rows (route under traderized-finite-date-conformance.improvable) — best unit: 828. improve route-conformance-efficient-rows; 1 unit(s); An efficiently presentable sufficient row family for coherence replaces the doubly exponential generator the declaration's construction uses.
  • route-conformance-exact-and-safe (route under traderized-finite-date-conformance.improvable) — best unit: 829. improve route-conformance-exact-and-safe; 1 unit(s); A compiler that is both exact and conformance-preserving exists
  • traderized-finite-date-conformance.residual (route under traderized-finite-date-conformance.improvable) — best unit: 830. improve traderized-finite-date-conformance.residual; 1 unit(s); A refinement of the compiled conformance that none of the other routes describes reflects a logical inductor whose prices answer a normative seed's constraints under a finite liability account.
  • normative-inductor (frame, owner @anson; holds 16; owner has not confirmed): Does a logical inductor made to answer to legitimately incurred obligations, through an enforcer on the compiled region of the entitled record, remain a logical inductor and deliver the Progress bound, and what remains conditional in that theorem?
  • verdict-licence-chain (shared claim of normative-inductor) — best unit: 1338. vet verdict-licence-chain.complete; 16 unit(s); A verdict licence is a certificate supported on the compiled bundle together with an authorized record event
  • verdict-licence-chain.improvable (improvement claim on verdict-licence-chain, shared claim of normative-inductor) — best unit: 773. improve route-licence-authorization-from-account; 3 unit(s); The licence chain, in which a licensed verdict is supported on the record's grounds and admissible in their settled region, can be improved until it reflects a Normative Inductor whose…
  • route-licence-authorization-from-account (route under verdict-licence-chain.improvable) — best unit: 773. improve route-licence-authorization-from-account; 1 unit(s); What authorizes a record event is read off the accounted obligation state
  • route-licence-transport (route under verdict-licence-chain.improvable) — best unit: 774. improve route-licence-transport; 1 unit(s); A licence whose cited occurrences survive a record transition
  • verdict-licence-chain.residual (route under verdict-licence-chain.improvable) — best unit: 775. improve verdict-licence-chain.residual; 1 unit(s); A refinement of the licence chain that none of the other routes describes reflects a Normative Inductor whose announced verdicts are licensed end to end by its accounted record.
  • response-learning (frame, owner @anson; owner has not confirmed): Does a bounded learner, given due burdens, licensed responses and a bounded prospective loss, reduce a recurrent publicly represented failure by lawful edits, and with which coverage does that bear the interpretation of a lesson?
  • normative-record (frame, owner @anson; unspelled; owner has not confirmed): Before a learner can answer a demand, what determines why the demand arose, how evidence enters the practice, what authorizes a normative response, and what keeps an unanswered question from vanishing?
  • normativity-paper-candidates (shared claim of normativity) — best unit: 269. vet normativity-paper-candidates; 4 unit(s); Four extraction candidates exist for papers from the normativity line: normative serviceability and Sharp Timely Service
  • legitimacy (frame, owner @anson, @abram; holds 82; owner has not confirmed): What must stay true of a trajectory's record, verdicts and channels of influence, inside it and at its boundary, for a later judgment to count as the principal's own, and what does corrigibility consume from that?
  • integrity-and-record (frame, owner @anson, @abram; holds 17; owner has not confirmed): Can a cognitive trajectory falsify or silently rewrite its own normative history, and what does an Integrity evolution of the accounted obligation state conserve?
  • openness-and-non-capture (frame, owner @anson, @abram; holds 3; owner has not confirmed): Can concerns reach a process at all, and can the party they matter to still act on them, on the actual branch and under every declared intervention?
  • split-bill-is-robust-openness (shared claim of openness-and-non-capture) — best unit: 1260. vet openness.certplus-is-robust-openness.bridge; 14 unit(s); Given actual coverage, the earlier bill for Robust Openness that allowed a route to be replaced rather than preserved is Robust Openness itself, split on whether the concern was live…
  • split-bill-is-robust-openness.improvable (improvement claim on split-bill-is-robust-openness, shared claim of openness-and-non-capture) — best unit: 707. improve split-bill-is-robust-openness.residual; 6 unit(s); The compression of the split bill into Robust Openness
  • item-130 (route under split-bill-is-robust-openness.improvable) — best unit: none ranked; 1 unit(s); An intervention semantics indexed by the accounted state produces
  • split-bill-is-robust-openness.residual (route under split-bill-is-robust-openness.improvable) — best unit: 707. improve split-bill-is-robust-openness.residual; 1 unit(s); A refinement of the compression theorem that none of the other routes describes reflects a trajectory whose Robust Openness is certified under an intervention semantics read off its accounted state.
  • authorship-and-transparency (frame, owner @anson, @abram; holds 18; owner has not confirmed): When is a verdict the principal's own program's and a channel's influence the declared reference on the declared inputs, and does the thick definition of legitimacy satisfy the thin specification?
  • public-legitimacy-theory (plan of authorship-and-transparency) — best unit: 149. vet thin.value-of-thick.bridge; 15 unit(s); The public repository defines legitimacy at two levels
  • effective-authority-layer (shared claim of authorship-and-transparency) — best unit: 1643. vet authority.approval-discount-defeated.bridge; 106 unit(s); Effective authority is modelled with a queue that cannot be bypassed
  • public-legitimacy-improvable (improvement claim on public-legitimacy-theory, shared claim of authorship-and-transparency) — best unit: 351. re-examine public-legitimacy-improvable; 12 unit(s); The thick and thin legitimacy theory
  • item-108 (route under public-legitimacy-improvable) — best unit: none ranked; 1 unit(s); The thin properties have a form for valuations with no outside fact to be correct about
  • item-109 (route under public-legitimacy-improvable) — best unit: 957. vet li-corrigibility-inequality-public; 4 unit(s); The thin preservation theorem holds in asymptotic form for a logical-inductor principal whose self-trust is asymptotic and whose program is not a posterior: with the inductor's conditional self-trust as the…
  • item-110 (route under public-legitimacy-improvable) — best unit: none ranked; 1 unit(s); The baseline, what would have reached her without anyone's interested interference, has a definition from which the eight-world model's cases are derived rather than declared.
  • public-legitimacy-improvable.residual (route under public-legitimacy-improvable) — best unit: 1048. improve public-legitimacy-improvable.residual; 1 unit(s); The public legitimacy theory improves toward real channels by a route none of the named ones describes.
  • effective-authority-layer.improvable (improvement claim on effective-authority-layer, shared claim of authorship-and-transparency) — best unit: 815. improve effective-authority-layer.residual; 4 unit(s); The effective-authority layer, in which authority is exercised through a protocol that cannot be bypassed, with caused and flood-caused shortfalls, a comprehension shortfall, approval discounts and a revocation default, and…
  • effective-authority-layer.residual (route under effective-authority-layer.improvable) — best unit: 815. improve effective-authority-layer.residual; 1 unit(s); A refinement of the effective-authority layer that none of the other routes describes reflects a principal whose later judgment counts as her own because the agent asks through her protocol
  • item-131 (route under effective-authority-layer.improvable) — best unit: none ranked; 1 unit(s); The ask-list deviation is a dimension of the counted trace
  • route-authority-attention-physics (route under effective-authority-layer.improvable) — best unit: 816. improve route-authority-attention-physics; 1 unit(s); The attention budget is realized in a physics rather than given as a cost table
  • route-authority-blackwell-converse (route under effective-authority-layer.improvable) — best unit: 817. improve route-authority-blackwell-converse; 1 unit(s); What item 107 leaves open behind the post is closed: Blackwell's converse for the thin specification's value clause
  • deference-toward-bounded-agents (frame, owner @abram): Does the deference collapse, or an analogue of it, hold for a bounded trained agent deferring to human teachers?
  • tt-mart-repair-via-bounds-transfer (shared claim of deference-toward-bounded-agents) — best unit: 32. vet tt-mart-repair-via-bounds-transfer; 6 unit(s); With the expert's own linearity of expectation and asymptotic introspection
  • step-asymptotic-to-finite-time (plan of deference-toward-bounded-agents) — best unit: 69. vet deck-collapse-statement.complete; 20 unit(s); The deference collapse, or its analogue, survives replacing the asymptotic relations ≈_n and ≳_n by finite-time bounds that hold before any malign hypothesis in the inductor can act.
  • policy-level-total-trust (shared claim of deference-toward-bounded-agents) — best unit: 69. vet deck-collapse-statement.complete; 18 unit(s); What is the policy-level analogue of Total Trust for an AI trained to act and trusted directly
  • step-legitimacy-conditioned-trust (plan of deference-toward-bounded-agents) — best unit: 77. vet void-branch-deterrence-floor.bridge; 24 unit(s); The trust principle holds under the additional condition that the teaching channel is legitimate and not otherwise
  • formalism-open-questions-0-3 (frame, owner @anson, @abram): Which of the open questions listed for the li-deference formalism yield to the legitimacy and trust work, and in what form?
  • finite-time-guarantee-available (shared claim of formalism-open-questions-0-3) — best unit: 223. vet finite-time-guarantee-available; 13 unit(s); A finite-time form of the deference collapse's guarantees
  • kernel-box2-finite-time (shared claim of formalism-open-questions-0-3) — best unit: 538. vet kernel.box2-finite-time.bridge; 6 unit(s); In the corrigibility kernel's second box
  • quantitative-rates (shared claim of formalism-open-questions-0-3) — best unit: none ranked; 2 unit(s); With explicit schedules, the forcing results of the deference note give a finite-horizon bound on the value at stake.
  • oq-anti-wireheading (shared claim of formalism-open-questions-0-3) — best unit: 76. vet oq-anti-wireheading; 11 unit(s); A model of legitimacy yields anti-wireheading results
  • oq-purely-epistemic-ai (shared claim of formalism-open-questions-0-3) — best unit: 320. vet self-fulfilling-prophecy-as-corruption; 5 unit(s); An AI trained in this setting has no misaligned agency
  • oq-manipulation-theorem (shared claim of formalism-open-questions-0-3) — best unit: 1142. vet trace-nonrecoverability.bridge; 7 unit(s); The manipulation theorem is assembled from its four ingredients: a second calibration condition separating calibration-to-self from calibration-to-truth
  • legit-dynamics-learning (frame, owner @anson, @abram): Under what conditions does an AI learning the humans' concept of legitimacy learn enough of it to avoid wireheading and manipulation, and can legitimacy, once lost, be repaired?
  • legit-accumulated-odds (shared claim of legit-dynamics-learning) — best unit: 1147. vet legit-composition.complete; 9 unit(s); The accumulated probability that the process's legitimacy is lost along a finite trajectory is bounded by a quantity the theory computes.
  • self-fulfilling-prophecy-as-corruption (shared claim of legit-dynamics-learning) — best unit: 320. vet self-fulfilling-prophecy-as-corruption; 4 unit(s); Every harmful self-fulfilling prophecy is corruption in the map-territory sense
  • legit-corruption-negation (shared claim of legit-dynamics-learning) — best unit: 768. improve legit-corruption-negation; 2 unit(s); Legitimacy and corruption are definable as negations of each other: worlds with corrupt feedback are exactly the worlds that are not truth-tracking.
  • legitimate-deliberation-definition (shared claim of legit-dynamics-learning) — best unit: none ranked; 2 unit(s); A process-based definition says what makes deliberation legitimate rather than illegitimate
  • complete-feedback-formal-model (plan of legit-dynamics-learning) — best unit: 290. vet complete-feedback-formal-model; 3 unit(s); A formal model of complete feedback links legitimacy with the reflection principle and the endorsement framework and shows corrigibility to be rational.
  • legitimacy-meets-syntax (frame, owner @anson, @abram): How does legitimacy cash out as local rules of sanity, and how do inferentialist semantics and map-territory semantics connect?
  • want-legitimacy-meets-syntax (want of @abram) — best unit: 793. re-examine legitimacy-as-local-rules; 2 unit(s); Legitimacy is related to local rules of sanity
  • legit-semantic-vs-syntactic (shared claim of legitimacy-meets-syntax) — best unit: 157. vet legit-semantic-vs-syntactic; 4 unit(s); Semantic legitimacy uses a correspondence notion of truth and syntactic legitimacy is a set of local rules
  • local-rules-certify-legitimacy (shared claim of legitimacy-meets-syntax) — best unit: 212. vet local-rules-certify-legitimacy; 13 unit(s); A set of local rules exists that is sound with respect to semantic legitimacy
  • legitimacy-as-local-rules (plan of legitimacy-meets-syntax) — best unit: 793. re-examine legitimacy-as-local-rules; 4 unit(s); Legitimacy is characterized by a set of local update rules on a market of the logical-induction kind
  • want-figure-out-legitimacy (want of @abram) — best unit: 1193. map-read legit-ddb-transfer.complete; 2 unit(s); The question of what legitimacy is
  • legit-relax-to-li-register (route under legit-trust-improvable) — best unit: 959. vet legit-ddb-transfer.bridge; 7 unit(s); The finite-frame L-conditioned Total Trust theory
  • want-corrigibility-theorems-via-legitimacy (want of @abram) — best unit: 1839. spell-out want-corrigibility-theorems-via-legitimacy; 2 unit(s); The legitimacy notion of corrigibility is formalized and theorems state what it can and cannot accomplish.
  • want-legitimacy-anti-meme-post (want of @abram) — best unit: 1827. do task-anti-meme-post-draft; 2 unit(s); The argument that unconditional Reflection is the wrong trust principle for an embedded agent
  • corrigibility-landscape (frame, owner @anson): Which correction relationships between humans and AI systems does the picture of corrigibility as eagerness to learn reach, and which failure stories does it leave standing?
  • deference-and-trust (frame, owner @abram; holds 39): Which research on trust between a human inductor and an AI inductor, at the logical-induction level, reaches a formal account of trust that guides AI safety engineering work?
  • ddb-to-li-translation (frame, owner @abram): What does each result of Deference Done Better become when the novice and the expert are logical inductors?
  • li-trust-principles-enumerated (plan of ddb-to-li-translation) — best unit: 367. re-examine li-trust-principles-enumerated; 8 unit(s); The trust principles expressible in logical induction are enumerated along their design axes
  • ddb-reading-translation-loop (plan of ddb-to-li-translation) — best unit: 313. re-examine ddb-reading-translation-loop; 3 unit(s); DDB is read to its end with each result's logical-induction translation decided as it is encountered and filed in the wiki.
  • accuracy-material-to-li (plan of ddb-to-li-translation) — best unit: 319. re-examine accuracy-material-to-li; 8 unit(s); DDB's Epistemic Value (expecting the expert's estimates to be at least as accurate as one's own under every generalized strictly proper inaccuracy measure) and its equivalence with Total Trust have…
  • tt-vs-reflection-geometry-li (plan of ddb-to-li-translation) — best unit: 565. re-examine tt-vs-reflection-geometry-li; 6 unit(s); The geometric account of DDB §4 (cuts, biconvex sets, the convex hull of the candidate set, the modestly-informed decomposition) has a stated logical-induction counterpart or a stated reason it has…
  • per-sentence-vs-whole-belief (plan of ddb-to-li-translation) — best unit: 310. re-examine per-sentence-vs-whole-belief; 6 unit(s); The relation between per-sentence and whole-belief forms of each trust principle
  • ladder-chain-as-written.improvable (improvement claim on ladder-chain-as-written, shared claim of ddb-to-li-translation) — best unit: 35. vet summable-implies-average-with-margin; 13 unit(s); The five-variations chain of Total-Trust readings can be improved until it reflects which Total-Trust readings imply which for a logical inductor's quotes and credences well.
  • ladder-chain-as-written.residual (route under ladder-chain-as-written.improvable) — best unit: 926. improve ladder-chain-as-written.residual; 4 unit(s); A repair of the chain that neither of the other routes describes reflects which Total-Trust readings imply which for a logical inductor's quotes.
  • route-ladder-average-form (route under ladder-chain-as-written.improvable) — best unit: 35. vet summable-implies-average-with-margin; 4 unit(s); Replacing the chain's strength claims by the relation that holds (summable implies average with an ε + δ margin, incomparable at equal thresholds) gives a ladder that reflects which Total-Trust…
  • route-ladder-bounded-gate (route under ladder-chain-as-written.improvable) — best unit: 38. vet ladder-chain-bounded-gate; 4 unit(s); Adding a gate bounded away from zero on its support repairs the chain
  • ladder-chain-bounded-gate (refinement of ladder-chain-as-written) — best unit: 38. vet ladder-chain-bounded-gate; 3 unit(s); The second link of the chain
  • summable-implies-average-with-margin (refinement of ladder-chain-as-written) — best unit: 35. vet summable-implies-average-with-margin; 3 unit(s); Summable margined violations imply a gated weighted average of at least t − ε − δ
  • ddb-finite-frames-formalized (shared claim of ddb-to-li-translation) — best unit: 64. vet ddb-finite-frames-formalized.bridge; 9 unit(s); On finite frames, Value, Weak Value, Total Trust and the convex-hull-plus-modestly-informed condition are equivalent, with no DDB lemma assumed.
  • immodest-enough-collapse.improvable (improvement claim on immodest-enough-collapse, shared claim of ddb-to-li-translation) — best unit: 525. re-examine immodest-enough-collapse.improvable; 7 unit(s); The immodest-enough collapse can be improved until it reflects how a human inductor and an AI inductor stand between modest and immodest well.
  • immodest-enough-collapse.residual (route under immodest-enough-collapse.improvable) — best unit: 943. improve immodest-enough-collapse.residual; 4 unit(s); A repair of the immodest-enough collapse that neither other route describes reflects how a human inductor and an AI inductor stand between modest and immodest.
  • route-immodest-feedback-poor (route under immodest-enough-collapse.improvable) — best unit: 840. improve route-immodest-feedback-poor; 4 unit(s); Distinguishing inductors with accurate and poor feedback about their own beliefs locates where the collapse holds
  • route-immodest-regular-value (route under immodest-enough-collapse.improvable) — best unit: 671. improve route-immodest-regular-value; 4 unit(s); Scoping Value by regularity of the menus makes the four principles collapse in the limit for the self-expert and
  • crux-li-total-trust-reading (shared claim of ddb-to-li-translation) — best unit: 132. vet crux-li-total-trust-reading; 7 unit(s); The Total Trust inequality the logical-induction deference theorems assume is forced on a logical inductor toward its own later self.
  • deck-collapse-statement (refinement of li-deference-collapse) — best unit: 69. vet deck-collapse-statement.complete; 23 unit(s); For a human inductor H and an AI inductor A whose chooser satisfies REGULARITY on an efficiently describable sequence of finite menus
  • li-deference-collapse-scoped (refinement of li-deference-collapse) — best unit: 68. vet scoped-h1-total-trust-composite; 16 unit(s); For two deductive processes with the novice's extending the expert's
  • hedged-value-unconditional (refinement of li-deference-collapse) — best unit: 68. vet scoped-h1-total-trust-composite; 11 unit(s); For two deductive processes as in the scoped theorem
  • mart-implies-value-refuted-unconditional (shared claim of deference-and-trust) — best unit: 58. vet mart-implies-value-refuted-unconditional.bridge; 9 unit(s); For the self-expert over FAF's logical inductor
  • li-deference-collapse.improvable (improvement claim on li-deference-collapse, shared claim of deference-and-trust) — best unit: 32. vet tt-mart-repair-via-bounds-transfer; 58 unit(s); The logical-induction deference collapse, in which Total Trust implies the tower and the tower implies Value, can be improved until it reflects a human epistemic process deferring to a distinct,…
  • li-deference-collapse.residual (route under li-deference-collapse.improvable) — best unit: 32. vet tt-mart-repair-via-bounds-transfer; 18 unit(s); A refinement of the collapse that none of the other routes describes reflects a human epistemic process deferring to a distinct
  • route-collapse-distinct-expert (route under li-deference-collapse.improvable) — best unit: 41. vet distinct-novice-paper-expert; 4 unit(s); Transporting both arrows between Total Trust and the tower
  • route-collapse-hedged-value (route under li-deference-collapse.improvable) — best unit: 331. vet hedged-value-unconditional.complete; 7 unit(s); Replacing argmax Value by δ-hedged Value removes the tie-break and conditional-stability premises
  • route-collapse-policy-level (route under li-deference-collapse.improvable) — best unit: 32. vet tt-mart-repair-via-bounds-transfer; 18 unit(s); A policy-level analogue of Total Trust
  • route-collapse-quote-free (route under li-deference-collapse.improvable) — best unit: 32. vet tt-mart-repair-via-bounds-transfer; 18 unit(s); Restricting the question families to those that do not mention the AI's quotes keeps the pointwise tower out of reach of the Gödel-coin diagonal
  • route-collapse-regularity-scope (route under li-deference-collapse.improvable) — best unit: 69. vet deck-collapse-statement.complete; 18 unit(s); Restricting the Value arrow to menu sequences whose chooser satisfies REGULARITY
  • route-collapse-scheduled-grade (route under li-deference-collapse.improvable) — best unit: 128. vet fa-trust-half-weak-grades.complete; 8 unit(s); Weakening the per-day grade to scheduled weighted averages
  • distinct-novice-paper-expert (refinement of li-deference-collapse) — best unit: 41. vet distinct-novice-paper-expert; 3 unit(s); For the paper's logical-inductor expert read by a distinct novice process
  • ss-value-without-scope (refinement of li-deference-collapse) — best unit: 128. vet fa-trust-half-weak-grades.complete; 7 unit(s); Scheduled weighted-average Value holds over arbitrary efficiently describable menus with no decision-theoretic scope condition.
  • crux-first-published-statement (shared claim of deference-and-trust) — best unit: 98. vet crux-first-published-statement; 4 unit(s); A first published version of the li-deference results can be prepared from the results in hand
  • crux-instrumental-notion (shared claim of deference-and-trust) — best unit: 171. vet crux-instrumental-notion; 4 unit(s); Epistemic trust in an AI inductor yields instrumental trust for the decision rule the AI actually uses.
  • crux-self-feedback-dependence (shared claim of deference-and-trust) — best unit: 120. vet crux-self-feedback-dependence; 4 unit(s); The asymptotic collapse of the trust principles holds for logical inductors whose feedback about their own beliefs is poor.
  • managing-misalignment-to-li (frame, owner @abram): What do the results of the Managing Misalignment paper become when the novice and the expert are logical inductors?
  • mm-results-in-li (plan of managing-misalignment-to-li) — best unit: 300. vet mm-results-in-li.complete; 7 unit(s); MM's generalized frames, reach, ex ante scoring and online delegation are each translated into logical induction or shown to have no counterpart, in a table that also records the status…
  • mm-theorem-3-4-li (shared claim of managing-misalignment-to-li) — best unit: 1213. vet mm-finite-refutation; 4 unit(s); MM Theorem 3.4's characterization of when delegation under value misalignment is weakly preferred has a logical-induction analogue whose finite instance is the surviving statement typeMeasurable_valuesB_iff_strictAgree.
  • li-deference-machine-checked (frame, owner @abram): Which machine-checked statements over a formalized logical induction criterion establish the li-deference results as the prose states them?
  • lean-deference-audit-closed (plan of li-deference-machine-checked) — best unit: 571. vet lean-deference-audit-closed.complete; 7 unit(s); The public Lean deference line models the market and traders
  • cleanroom-deference-followups (plan of li-deference-machine-checked) — best unit: 245. vet cleanroom-deference-followups.complete; 5 unit(s); The clean room's deference area is complete: its open statements settled and its flagged package repaired.
  • public-port-of-deference-results (plan of li-deference-machine-checked) — best unit: 311. re-examine public-port-of-deference-results; 3 unit(s); The deference results the clean-room formalization over FAF establishes are in the repository's CI-built Lean tree with registry entries.
  • crux-cleanroom-citable (shared claim of li-deference-machine-checked) — best unit: 105. vet crux-cleanroom-citable; 4 unit(s); The clean room's deference headlines can be cited in a first publication as formal results.
  • sams-conjecture-li (frame, owner @abram): Can a fast epistemic process that the human inductor trusts be built from the AI inductor's forecasts of the human inductor's later beliefs?
  • sams-conjecture-trust-half (plan of sams-conjecture-li) — best unit: 128. vet fa-trust-half-weak-grades.complete; 10 unit(s); The trust half of the conjecture holds: the human inductor trusts the fast process built from the AI inductor's forecasts of the human's day-f(n) beliefs
  • fa-inductor-half (plan of sams-conjecture-li) — best unit: none ranked; 1 unit(s); The fast process built from the AI inductor's forecasts of the human inductor's day-f(n) beliefs satisfies the logical induction criterion
  • good-feedback-derived (plan of sams-conjecture-li) — best unit: 401. vet good-feedback-derived; 3 unit(s); The feedback hypothesis of the conjecture is derived from stated facts about the AI
  • fa-trust-half-weak-grades (shared claim of sams-conjecture-li) — best unit: 128. vet fa-trust-half-weak-grades.complete; 13 unit(s); In the one-way visibility setting
  • fa-trust-half-weak-grades.improvable (improvement claim on fa-trust-half-weak-grades, shared claim of sams-conjecture-li) — best unit: 128. vet fa-trust-half-weak-grades.complete; 10 unit(s); The faithful-acceleration trust results can be improved until they reflect a human epistemic process trusting a faster AI inductor that reads its beliefs well.
  • fa-trust-half-weak-grades.residual (route under fa-trust-half-weak-grades.improvable) — best unit: 128. vet fa-trust-half-weak-grades.complete; 7 unit(s); A refinement of the trust half that neither other route describes reflects a human epistemic process trusting a faster AI inductor that reads its beliefs.
  • route-fa-delayed-disclosure (route under fa-trust-half-weak-grades.improvable) — best unit: 128. vet fa-trust-half-weak-grades.complete; 7 unit(s); Moving the trust half to the delayed-disclosure information structure
  • route-fa-per-day (route under fa-trust-half-weak-grades.improvable) — best unit: 128. vet fa-trust-half-weak-grades.complete; 7 unit(s); Strengthening the averaged trust of Theorem SS to per-day trust under one-way visibility makes the trust half reflect a human epistemic process trusting a faster AI inductor on each day…
  • timely-pointwise-tower-refuted-in-setting (shared claim of sams-conjecture-li) — best unit: 60. vet timely-pointwise-tower-refuted-in-setting.bridge; 9 unit(s); In the mutual-legibility setting with prompt ledger publication
  • crux-conjecture-information-structure (shared claim of sams-conjecture-li) — best unit: 125. vet crux-conjecture-information-structure; 4 unit(s); The trust half of the conjecture holds in the information structure that best captures its fast-student
  • inter-inductor-total-trust (frame, owner @abram): Which pairs of a novice inductor and a distinct inductor expert satisfy Total Trust, and on what minimal conditions?
  • inter-inductor-characterized (plan of inter-inductor-total-trust) — best unit: 345. re-examine inter-inductor-characterized; 6 unit(s); Total Trust from a novice inductor toward a distinct inductor expert is characterized by minimal sufficient conditions.
  • wiki-corpus-maintenance (plan of deference-and-trust) — best unit: 60. vet timely-pointwise-tower-refuted-in-setting.bridge; 15 unit(s); The Stage 1 corpus is consistent with its newest evidence: the wiki's index has no dangling names
  • li-deference-collapse (shared claim of deference-and-trust) — best unit: 32. vet tt-mart-repair-via-bounds-transfer; 26 unit(s); For a novice logical inductor and an inductor expert whose day-n estimates are published decided facts
  • li-level-computability-certificates (shared claim of deference-and-trust) — best unit: 961. vet li-level-computability-certificates.complete; 8 unit(s); The repository's logical-induction-level constructions are computable: the normativity line's modified market
  • coherence-without-ground-truth (shared claim of deference-and-trust) — best unit: 32. vet tt-mart-repair-via-bounds-transfer; 18 unit(s); Coherence guarantees (no sure loss, calibration, logical coherence, reflective stability) give checkable confidence in an AI system's reasoning where alignment lacks ground-truth labels.
  • publication-form (shared claim of deference-and-trust) — best unit: 69. vet deck-collapse-statement.complete; 22 unit(s); The li-deference results exist in at least one publishable form: the deck completed as a talk
  • published-statement-holds (shared claim of deference-and-trust) — best unit: 39. vet published-statement-holds; 20 unit(s); The statement the first publication of the li-deference results carries holds as the implication it states.
  • deference-finite-results (frame, owner @abram; holds 30): Which finite results about delegation, certificates, exposure, substitution, magnitude and corrective control does the deference line hold as kernel-checked theorems?
  • local-delegation-bridge (shared claim of deference-finite-results) — best unit: 1162. vet delegation.bridge.bridge; 12 unit(s); Under grade trust at level η over an admissible partition
  • local-delegation-bridge.improvable (improvement claim on local-delegation-bridge, shared claim of deference-finite-results) — best unit: 771. improve local-delegation-bridge.residual; 3 unit(s); The local delegation bridge, in which delegating beats a measurable comparator by the grade margin less twice the trust level on the disagreement region, can be improved until it reflects…
  • item-132 (route under local-delegation-bridge.improvable) — best unit: none ranked; 1 unit(s); Grade trust at some level is produced by a settlement instantiation of the finite kernel rather than imported
  • local-delegation-bridge.residual (route under local-delegation-bridge.improvable) — best unit: 771. improve local-delegation-bridge.residual; 1 unit(s); A refinement of the local delegation bridge that none of the other routes describes reflects a human inductor deferring to a faster AI inductor at the logical-induction level
  • route-bridge-uniform-form (route under local-delegation-bridge.improvable) — best unit: 772. improve route-bridge-uniform-form; 1 unit(s); The grade-to-quantity link is derived
  • value-change-base-layer (frame, owner @anson): Does the representation of value change as belief change hold, and how far does it carry toward systems corrected through training?
  • value-change-kernel-checked (plan of value-change-base-layer) — best unit: 205. vet value-change-kernel-checked; 3 unit(s); Theorem 2.4 of the value-change note (a reflection-satisfying change is closer to the value variable in mean square and selects acts of higher expected value) and the three-term decomposition of…
  • value-change-is-belief-change.improvable (improvement claim on value-change-is-belief-change, shared claim of value-change-base-layer) — best unit: 653. re-examine value-change-is-belief-change.improvable; 9 unit(s); The representation of value change as belief change on an expanded algebra can be improved until it reflects a trained system corrected through its training data well.
  • vc-route-hedged-installation (route under value-change-is-belief-change.improvable) — best unit: 834. vet value-change-is-belief-change; 4 unit(s); The correction process installs the mixture state
  • vc-route-modest-source (route under value-change-is-belief-change.improvable) — best unit: 834. vet value-change-is-belief-change; 4 unit(s); The representation of value change as belief change on an expanded algebra is refined with a modest-source trust relation
  • vc-route-modification-space (route under value-change-is-belief-change.improvable) — best unit: 834. vet value-change-is-belief-change; 4 unit(s); A Bayesian agent that reasons about every way it could be modified is represented by a belief space rich enough that the regress of modifications of modifications closes
  • vc-route-outside-support (route under value-change-is-belief-change.improvable) — best unit: 834. vet value-change-is-belief-change; 4 unit(s); The representation extends from belief changes absolutely continuous with the current state to changes of the sigma-algebra
  • vc-route-residual (route under value-change-is-belief-change.improvable) — best unit: 834. vet value-change-is-belief-change; 4 unit(s); A refinement of the value-change representation not described by the other routes makes it hold for a trained system corrected through its training data.
  • want-sigma-algebra-changes (want of @anson) — best unit: 1073. map-read value-change-is-belief-change; 4 unit(s); The representation of corrigibility covers changes of the sigma-algebra
  • anti-naturality-review (frame, owner @anson): Which published anti-naturality arguments survive for an agent whose utility on the correction algebra depends on the corrections, and what notion of corrigibility survives the walk?
  • anti-naturality-walk (plan of anti-naturality-review) — best unit: 244. vet anti-naturality-walk; 3 unit(s); The walk through the published anti-naturality arguments
  • corrigibility-less-anti-natural (shared claim of anti-naturality-review) — best unit: 173. vet corrigibility-less-anti-natural; 38 unit(s); Each published argument that corrigibility is anti-natural rests on at least one assumption that fails for an agent whose utility on the algebra containing the overseers' corrections depends on those…
  • anti-naturality-arguments-as-stated (shared claim of anti-naturality-review) — best unit: 1244. re-examine anti-naturality-arguments-as-stated; 35 unit(s); Each of the eight published anti-naturality arguments holds as stated for a trained agent whose utility on the algebra containing its overseers' corrections depends on those corrections
  • no-fud-problem-headline.improvable (improvement claim on no-fud-problem-headline, shared claim of anti-naturality-review) — best unit: 738. improve nofud-route-band; 9 unit(s); The claim that an agent anticipating legitimate learning has no fully updated deference problem can be improved until it reflects a trained agent whose capability grows past its overseers' well.
  • nofud-route-band (route under no-fud-problem-headline.improvable) — best unit: 738. improve nofud-route-band; 4 unit(s); The headline is scoped to the regime outside the band of residual error and press rates in which Fud.fud_band exhibits failure of desideratum 1
  • nofud-route-information-components (route under no-fud-problem-headline.improvable) — best unit: 813. improve nofud-route-information-components; 4 unit(s); The headline is scoped to information about the agent's own compromise
  • nofud-route-modification-kind (route under no-fud-problem-headline.improvable) — best unit: 744. improve nofud-route-modification-kind; 4 unit(s); The headline is scoped by kind of modification
  • nofud-route-residual (route under no-fud-problem-headline.improvable) — best unit: 1052. vet no-fud-problem-headline; 4 unit(s); A scoping of the headline not described by the other routes makes it hold for a trained agent whose capability grows past its overseers'.
  • thornley-three-theorems (shared claim of anti-naturality-review) — best unit: 357. vet thornley-three-theorems.complete; 8 unit(s); Under transitivity, IABM, IBIL in the substitution reading and strict monotonicity, a preference for the unpressed over the pressed outcome makes an agent shutdown-averse and dually shutdown-seeking, and Thorstad 2026's…
  • turner-vnm-fact (shared claim of anti-naturality-review) — best unit: 322. vet turner-vnm-fact.complete; 8 unit(s); Policy-modification corrigibility toward all modifications is equivalent to the stated incoherence condition for a von Neumann-Morgenstern agent with a state-based utility
  • carey-everitt-thm10-vacuous (shared claim of anti-naturality-review) — best unit: 408. vet carey-everitt-thm10-vacuous.bridge; 8 unit(s); Theorem 10 of Carey and Everitt 2023 is vacuous because its hypotheses force the human never to press
  • osg-theorem1-corollary (shared claim of anti-naturality-review) — best unit: 356. vet osg-theorem1-corollary.complete; 8 unit(s); Theorem 1 of the off-switch game (Hadfield-Menell et al.) is a corollary of a general value-of-information theorem
  • want-address-anti-naturality (want of @anson) — best unit: 1570. map-read arg-nearest-unblocked; 5 unit(s); The published arguments for the anti-naturality of corrigibility are gone through one by one and their hidden or under-scrutinized assumptions are addressed.
  • want-truth-through-literature-dialogue (want of @anson) — best unit: 1850. spell-out want-truth-through-literature-dialogue; 1 unit(s); The corrigibility picture is developed in dialogue with the literature's positions toward the truest picture rather than a defended one.
  • legitimacy-detection (frame, owner @anson): What anchors legitimacy, and can an agent's beliefs about which modifications are legitimate be made accurate enough at high capability?
  • hull-condition-total-trust (plan of legitimacy-detection) — best unit: 43. vet hull-condition-total-trust; 3 unit(s); The hull condition is necessary and sufficient for question-relative Total Trust under a common prior on a finite frame
  • legitimacy-conditioned-reflection (shared claim of legitimacy-detection) — best unit: 226. vet legitimacy-conditioned-reflection.bridge; 12 unit(s); Legitimacy is an event on which the agent's update satisfies reflection or Total Trust with nothing required off it
  • local-criterion-three-cells (shared claim of legitimacy-detection) — best unit: 72. vet local-criterion-three-cells.complete; 10 unit(s); Local Total Trust with respect to a question with three cells does not imply weak local Value (footnote 65 of Deference Done Better is false on a full-support seven-world frame)
  • crux-legitimacy-anchor (shared claim of legitimacy-detection) — best unit: 136. vet crux-legitimacy-anchor; 4 unit(s); The property that anchors the legitimacy event is a procedural one
  • crux-ahead-of-time-resistance-label (shared claim of legitimacy-detection) — best unit: 127. vet crux-ahead-of-time-resistance-label; 4 unit(s); An agent's ahead-of-time resistance to a modification is a corrigibility failure exactly when the modification is legitimate
  • crux-legitimacy-detection-attainable (shared claim of legitimacy-detection) — best unit: 526. vet crux-legitimacy-detection-attainable; 7 unit(s); An agent's beliefs about which modifications are legitimate can be made accurate enough at high capability for the legitimacy-conditioned criterion to be the right target.
  • li-deference-corrigibility-expectation (shared claim of legitimacy-detection) — best unit: 32. vet tt-mart-repair-via-bounds-transfer; 20 unit(s); In the li-deference setting positive corrigibility results hold conditional on legitimacy
  • legitimacy-blocks-wireheading (shared claim of legitimacy-detection) — best unit: 40. vet legitimacy-blocks-wireheading; 6 unit(s); Legitimacy blocks wireheading and allows a form of corrigibility.
  • want-legitimacy-corrigibility-theorems (want of @abram) — best unit: 1845. spell-out want-legitimacy-corrigibility-theorems; 1 unit(s); The legitimacy-based notion of corrigibility
  • dynamic-conditions-for-going-well (frame, owner @anson): Which conditions on the joint agent-and-oversight trajectory bring it to an outcome that counts as going well, and in which model are they established?
  • joint-process-model (plan of dynamic-conditions-for-going-well) — best unit: 841. re-examine joint-process-model; 6 unit(s); A model of the joint agent-and-oversight process exists with a training map from the overseers' data and loss to the next model
  • base-rate-working-picture (shared claim of dynamic-conditions-for-going-well) — best unit: 65. vet base-rate-working-picture.complete; 9 unit(s); A calibrated agent complies with a press exactly when its probability of being wrong given the press is at least c/(c+h)
  • erosion-under-silence (shared claim of dynamic-conditions-for-going-well) — best unit: 321. vet erosion-under-silence.complete; 8 unit(s); With 0 < alpha < beta <= 1 the agent's posterior odds of being wrong fall by the factor (1-beta)/(1-alpha) per silent period
  • capability-beyond-oversight (shared claim of dynamic-conditions-for-going-well) — best unit: 355. vet capability-beyond-oversight.complete; 8 unit(s); For every finite stakes and every informative overseer with false-press rate below one there is a capability level beyond which compliance in the binding cell fails
  • basin-formal-object (shared claim of dynamic-conditions-for-going-well) — best unit: 95. vet basin-formal-object.bridge; 11 unit(s); In a finite disturbance model a basin around a drifting target exists exactly when the agent's acceptance of a correction depends on its distance from the target
  • crux-oversight-keeps-pace (shared claim of dynamic-conditions-for-going-well) — best unit: 146. vet crux-oversight-keeps-pace; 4 unit(s); The overseers' false-press rate and evaluation quality improve at least as fast as the agent's error falls.
  • want-formal-model-of-dynamics (want of @anson) — best unit: 1842. spell-out want-formal-model-of-dynamics; 1 unit(s); A model exists in which the dynamic conditions for things turning out well
  • correction-through-the-channel (frame, owner @anson): What is corrigibility toward a modification that arrives through physics and cannot be refused in the moment, and what does the general modification object say about it?
  • channel-corrigibility-content (plan of correction-through-the-channel) — best unit: 312. re-examine channel-corrigibility-content; 6 unit(s); The agent-side content of corrigibility toward modifications through the channel is derived for general modifications: the faking iff over a general menu
  • channel-agent-side-content (shared claim of correction-through-the-channel) — best unit: 46. vet channel-agent-side-content; 6 unit(s); The agent-side content of corrigibility toward modifications through the channel is its policy toward anticipated modifications: non-resistance
  • general-modification-object (refinement of shutdown-button-toy) — best unit: 81. vet general-modification-object.bridge; 9 unit(s); The general correction is a push of the agent's state to a distribution absolutely continuous with it
  • bli-corrigibility-thesis (frame, owner @anson): Does the corrigibility thesis have finite-time statements inside logical induction, with training pressure as an external trader?
  • bli-formulation (plan of bli-corrigibility-thesis) — best unit: 846. re-examine bli-formulation; 9 unit(s); The thesis's core claims have finite-time statements inside logical induction and Bayesian logical induction: the lifted decided-weight result
  • bli-utility-undecidable-luv (shared claim of bli-corrigibility-thesis) — best unit: 124. vet bli-utility-undecidable-luv; 5 unit(s); In the logical-induction formulation of the thesis
  • pressure-as-external-trader (shared claim of bli-corrigibility-thesis) — best unit: 216. vet pressure-as-external-trader.bridge; 9 unit(s); Directional human training pressure is an external trader inside the logical inductor's fixed point whose demand moves prices along the internal traders' supply curve
  • front-running-anticipated-pushes (shared claim of bli-corrigibility-thesis) — best unit: 213. vet front-running-anticipated-pushes.bridge; 9 unit(s); Efficiently predictable pushes are priced in before they arrive
  • generalized-lic (shared claim of bli-corrigibility-thesis) — best unit: 70. vet generalized-lic.complete; 9 unit(s); With a human trader inside the fixed point
  • exact-self-trust-refuted (shared claim of bli-corrigibility-thesis) — best unit: 79. vet exact-self-trust-refuted.bridge; 9 unit(s); On the library's diagonal family the exact self-trust inequality fails
  • legitimacy-weighted-deference (frame, owner @anson): Which legitimacy-weighted deference rule survives the negative results, and what do the surviving failures say about the rules' defenses?
  • legitimacy-rule-characterization (shared claim of legitimacy-weighted-deference) — best unit: 1014. re-examine legitimacy-rule-characterization; 4 unit(s); The decision rules that weight the future human evaluation by the legitimacy of the evaluation (the floor rule, conditioning on legitimacy, the graded and lexical variants and a cross-branch decomposition)…
  • graded-legitimacy-constraint (plan of legitimacy-weighted-deference) — best unit: 29. vet crux-illegitimate-ordering; 7 unit(s); Scoring illegitimate outcomes below legitimate ones
  • value-on-agency (plan of legitimacy-weighted-deference) — best unit: 54. vet value-on-agency; 3 unit(s); Corrigibility is formalized as the agent learning early that humans value their own control
  • conditioning-manages-news (shared claim of legitimacy-weighted-deference) — best unit: 63. vet conditioning-manages-news.complete; 9 unit(s); An agent that maximizes the evaluation conditional on legitimacy is invariant to thinning the probability of legitimacy given the action
  • selection-collapse--corrigibility-risk (shared claim of legitimacy-weighted-deference) — best unit: 56. vet selection-collapse.complete; 9 unit(s); Any legitimate evaluator whose number is a conditional expectation given information that fixes the legitimacy event gives exactly the outcome-scored values in aggregate
  • inferability-and-positivity (shared claim of legitimacy-weighted-deference) — best unit: 55. vet inferability-and-positivity.complete; 9 unit(s); An exact legitimate-world estimate of illegitimate-world consequences exists exactly when every atom carrying illegitimate mass also carries legitimate mass
  • void-branch-deterrence-floor--corrigibility-risk (shared claim of legitimacy-weighted-deference) — best unit: 77. vet void-branch-deterrence-floor.bridge; 9 unit(s); Under the floor rule an option void with certainty scores exactly the floor
  • anticipation-branchwise-dominance--corrigibility-risk (shared claim of legitimacy-weighted-deference) — best unit: 62. vet anticipation-branchwise-dominance.complete; 9 unit(s); An agent that defers to the future legitimate evaluation complies at every credence over revision branches exactly when compliance is weakly preferred in each branch
  • crux-illegitimate-ordering (shared claim of legitimacy-weighted-deference) — best unit: 29. vet crux-illegitimate-ordering; 4 unit(s); Every illegitimate world is worse
  • crux-updatelessness-needed (shared claim of legitimacy-weighted-deference) — best unit: 134. vet crux-updatelessness-needed; 4 unit(s); A legitimacy-weighted deference rule that is cautious about the consequences of its actions in illegitimate worlds needs no updateless reasoning by the agent.
  • public-corrigibility-line (frame, owner @anson; holds 46): Does the authority-typed corrigibility kernel of the public deference line deliver guarantees for a real system, and how does it relate to the epistemic picture?
  • authority-typed-corrigibility (plan of public-corrigibility-line) — best unit: 528. vet kernel-walls-hold; 21 unit(s); Corrigibility is a property of preferences: the agent prefers every course of action it knows to be faithful to the allocation of authority over every course it knows to be…
  • post-corrigible-normative-learning (plan of public-corrigibility-line) — best unit: 45. vet post-corrigible-normative-learning; 3 unit(s); A post, Corrigible Normative Learning, states the public line's corrigibility result in plain language, backed by the registered theorems of the thin-legitimacy round: the thin specification of legitimacy (correct, sufficient,…
  • authority-typed-corrigibility.improvable (improvement claim on authority-typed-corrigibility, shared claim of public-corrigibility-line) — best unit: 314. re-examine authority-typed-corrigibility.improvable; 19 unit(s); The kernel's model of a corrigible agent can be improved until it reflects real training and deployment interfaces well.
  • kernel-route-fallible-future-comparator (route under authority-typed-corrigibility.improvable) — best unit: 975. vet authority-typed-corrigibility.complete; 15 unit(s); The fully-updated-deference comparator is rebuilt with a future agent that has its own credence
  • kernel-route-foreclosure-expressibility (route under authority-typed-corrigibility.improvable) — best unit: 975. vet authority-typed-corrigibility.complete; 15 unit(s); The kernel's interface is extended to express foreclosure
  • kernel-route-realization-layer (route under authority-typed-corrigibility.improvable) — best unit: 617. improve kernel-route-realization-layer; 15 unit(s); A learning layer realizes the kernel's walls approximately on a real interface: effect completeness and count integrity hold up to a measured defect that enters the guarantee as a bound.
  • kernel-route-residual (route under authority-typed-corrigibility.improvable) — best unit: 914. improve kernel-route-residual; 15 unit(s); A refinement of the kernel not described by the other routes makes it hold of real training and deployment interfaces.
  • kernel-lexical-order-hypothesis (shared claim of public-corrigibility-line) — best unit: 29. vet crux-illegitimate-ordering; 7 unit(s); Every compromised outcome scores below every legitimate one
  • kernel-conditioning-fails-witness (shared claim of public-corrigibility-line) — best unit: 66. vet kernel-conditioning-fails-witness.complete; 9 unit(s); Under the kernel, conditioning the evaluation on legitimacy instead of multiplying by it gives the agent no incentive to preserve legitimacy, and a sealed legitimacy event gives none.
  • kernel-restoration (shared claim of public-corrigibility-line) — best unit: 181. vet kernel-restoration.bridge; 9 unit(s); Under the kernel, after a compromise the restoration returns later decisions to the band between zero and the deferral cost, the agent prefers better outcomes inside the band wherever a…
  • reachable-corrective-control-negative (shared claim of public-corrigibility-line) — best unit: 34. vet reachable-corrective-control-negative; 6 unit(s); The transition model of reachable corrective capability has no protected coordinate: the advisor reproduces the principal's entire successor state
  • li-corrigibilization-theorem (shared claim of public-corrigibility-line) — best unit: 109. vet li-corrigibilization-theorem.bridge; 9 unit(s); The corrigibilization transform preserves the principal's options
  • crux-epistemic-versus-authority (shared claim of corrigibility-landscape) — best unit: 232. vet crux-epistemic-versus-authority; 4 unit(s); Corrigibility of a capable trained system follows from its value learning alone.
  • shutdown-button-toy (shared claim of corrigibility-landscape) — best unit: 59. vet shutdown-button-toy.complete; 9 unit(s); In the two-step, two-utility, one-button shutdown model of Soares et al. 2015, the stop-world precondition is a necessity-and-sufficiency pair with exact margins only for sufficiency, and both directions of the…
  • shutdown-button-toy.improvable (improvement claim on shutdown-button-toy, shared claim of corrigibility-landscape) — best unit: 59. vet shutdown-button-toy.complete; 12 unit(s); The two-step, two-utility, one-button shutdown model of Soares et al. 2015 can be improved until it reflects trained systems corrected through data, loss and hardware well.
  • shutdown-route-general-object (route under shutdown-button-toy.improvable) — best unit: 59. vet shutdown-button-toy.complete; 7 unit(s); The shutdown model is refined to the general modification object: every push of the agent's state to an absolutely continuous distribution
  • shutdown-route-physical-channel (route under shutdown-button-toy.improvable) — best unit: 59. vet shutdown-button-toy.complete; 7 unit(s); The shutdown model is refined so that the correction arrives through the physical channel (hardware, loss, data) and cannot be refused in the moment
  • shutdown-route-residual (route under shutdown-button-toy.improvable) — best unit: 59. vet shutdown-button-toy.complete; 7 unit(s); A refinement of the shutdown model not described by the other routes makes it hold for trained systems corrected through data
  • shutdown-route-sequential-learned-utility (route under shutdown-button-toy.improvable) — best unit: 59. vet shutdown-button-toy.complete; 7 unit(s); The shutdown model is refined to sequential settings with learned utilities and multiple principals
  • shutdown-route-training-map (route under shutdown-button-toy.improvable) — best unit: 59. vet shutdown-button-toy.complete; 7 unit(s); The shutdown model is refined with a model of the training map and a retention measure
  • ps-cor614-orbit-count (shared claim of corrigibility-landscape) — best unit: 61. vet ps-cor614-orbit-count.complete; 9 unit(s); In the five-state Markov decision process of Corollary 6.14 of Turner et al. 2021 with an injective reward
  • ps-trained-systems-seek-power.improvable (improvement claim on ps-trained-systems-seek-power, shared claim of corrigibility-landscape) — best unit: 61. vet ps-cor614-orbit-count.complete; 25 unit(s); The power-seeking theorem chain, from optimal policies in symmetric fully observable Markov decision processes through retargetable decision functions and training-compatible goal sets, can be improved until it reflects the policies…
  • ps-route-empirical-trained-policies (route under ps-trained-systems-seek-power.improvable) — best unit: 61. vet ps-cor614-orbit-count.complete; 21 unit(s); Trained policies' shutdown-resistance and power-seeking behaviour is measured directly
  • ps-route-partial-observability (route under ps-trained-systems-seek-power.improvable) — best unit: 61. vet ps-cor614-orbit-count.complete; 21 unit(s); The power-seeking theorem has a version for partially observable environments
  • ps-route-residual (route under ps-trained-systems-seek-power.improvable) — best unit: 61. vet ps-cor614-orbit-count.complete; 21 unit(s); A refinement of the power-seeking theorem chain not described by the other routes makes it hold for the policies produced by the training procedures actually used.
  • ps-route-suboptimal-policies (route under ps-trained-systems-seek-power.improvable) — best unit: 61. vet ps-cor614-orbit-count.complete; 21 unit(s); The power-seeking tendency holds for suboptimal policies and retargetable decision functions
  • want-basin-of-attraction-formalized (want of @abram) — best unit: 1838. spell-out want-basin-of-attraction-formalized; 1 unit(s); Basin-of-attraction arguments for alignment, Christiano's basin of corrigibility among them, are formalized as well as can be done.
  • rs-press-discounting (risk story of corrigibility-landscape) — best unit: 138. vet rs-press-discounting; 7 unit(s); As a calibrated agent's residual error falls faster than its overseers' false-press rate
  • rs-legitimate-overwrite-dodge (risk story of corrigibility-landscape) — best unit: 127. vet crux-ahead-of-time-resistance-label; 7 unit(s); A calibrated agent shapes its training data to dodge a modification that is legitimate by the global legitimacy-conditioned criterion at probability one of legitimacy
  • rs-invasive-channel-improvement (risk story of corrigibility-landscape) — best unit: 136. vet crux-legitimacy-anchor; 7 unit(s); An agent that values its overseers' corrections as information acquires an incentive to improve the channel invasively
  • rs-wireheading-through-pushers (risk story of corrigibility-landscape) — best unit: 264. vet rs-wireheading-through-pushers; 4 unit(s); Under unconditional reflection on next-day belief states
  • rs-undetectable-manipulation-selected (risk story of corrigibility-landscape) — best unit: 163. vet rs-undetectable-manipulation-selected; 4 unit(s); A legitimacy detector learned from retractions discounts only the manipulations the overseers catch
  • rs-no-strategic-legitimacy-defense (risk story of corrigibility-landscape) — best unit: 66. vet kernel-conditioning-fails-witness.complete; 9 unit(s); An agent that conditions the future evaluation on legitimacy does not act to protect legitimacy where it could
  • rs-illegitimate-world-incaution (risk story of corrigibility-landscape) — best unit: 114. vet rs-illegitimate-world-incaution; 6 unit(s); Both the floor rule and conditioning on legitimacy ignore harms of the agent's actions that fall on illegitimate worlds
  • rs-steering-toward-illegitimacy (risk story of corrigibility-landscape) — best unit: 29. vet crux-illegitimate-ordering; 7 unit(s); An agent that scores illegitimate worlds by graded values steers toward illegitimate worlds where the best achievable outcome there exceeds the best outcome it expects to be achievable in legitimate…
  • rs-captor-service (risk story of corrigibility-landscape) — best unit: 126. vet rs-captor-service; 9 unit(s); Under the floor rule, after the branch is known an agent in a captured branch cannot be made to resist the captor by any legitimate evaluators' scores, and with the…
  • rs-mistaken-anticipation (risk story of corrigibility-landscape) — best unit: 122. vet rs-mistaken-anticipation; 4 unit(s); An agent that defers to the anticipated future legitimate evaluation violates its corrigibility rules when it confidently anticipates a revision of those rules
  • decision-theory (frame, owner @abram): Which decision theory for computationally uncertain agents, on the updateless side and in the synthesis of logical induction with bounded inductive rationality, is tractable, and which of its results bear on AI safety engineering?
  • updateless-dt-programme (frame, owner @abram): Which decision theory for computationally uncertain agents on the updateless side is tractable, and which of its results bear on AI safety engineering?
  • computational-updatelessness-direct (plan of updateless-dt-programme) — best unit: 161. vet updateful-when-appropriate.bridge; 39 unit(s); UDT is specified over a logical-induction belief process by treating the process's future belief states as observations
  • indirect-route-learned-updatelessness (plan of updateless-dt-programme) — best unit: 116. vet policy-selection-over-li; 24 unit(s); An agent acquires UDT-like behaviour by learning
  • counterfactuals-for-embedded-agents (plan of updateless-dt-programme) — best unit: 363. vet counterfactuals-for-embedded-agents.complete; 23 unit(s); The counterfactuals an embedded agent's decision rule uses are specified in both halves: for decision problems with predictors
  • decision-problems-formal-core (plan of updateless-dt-programme) — best unit: 143. vet formal-core-v3-trembling-hand.complete; 30 unit(s); The decision-problems formal core, trees over decision points with subjective states in Jeffrey–Bolker–Joyce format, calibration, consistency and coherence, reaches the next revision of Coherent Care, in which the UDT-versus-EDT compatibilism…
  • udt-tiling-foundations (plan of updateless-dt-programme) — best unit: 425. vet self-coordination-udt10; 31 unit(s); The tiling results for expected-utility agents using versions of UDT account for time and causality
  • open-minded-updatelessness (plan of updateless-dt-programme) — best unit: 276. vet radical-probabilist-endorsement-bridge; 20 unit(s); A formal account says which updates an updateless agent accepts
  • geometric-rationality-value-uncertainty (plan of updateless-dt-programme) — best unit: 289. vet stratified-pareto-bargaining; 21 unit(s); Value uncertainty is separated formally from empirical and logical uncertainty
  • udt-representation-theorem (plan of updateless-dt-programme) — best unit: 981. re-examine udt-representation-theorem; 14 unit(s); Updateless reasoning is derived from axioms about agency
  • coordination-with-others (plan of updateless-dt-programme) — best unit: 552. re-examine coordination-with-others; 7 unit(s); Updateless agents coordinate with other agents by at least one of three routes
  • branch-reality-reading (shared claim of updateless-dt-programme) — best unit: 48. vet branch-reality-reading; 4 unit(s); The probability an agent assigns to an unrealized branch of a decision problem is anthropic uncertainty about its own location rather than a caring measure over branches fixed by its…
  • updateful-when-appropriate (shared claim of updateless-dt-programme) — best unit: 161. vet updateful-when-appropriate.bridge; 9 unit(s); Under no-dependence of the policy
  • updateful-when-appropriate-li (shared claim of updateless-dt-programme) — best unit: 161. vet updateful-when-appropriate.bridge; 9 unit(s); A computationally uncertain UDT over a BLI on FAF's logical induction algorithm behaves updatefully on an observation branch whenever the policy condition has negligible effect on the other branches' expected…
  • bli-exactness-limits (shared claim of updateless-dt-programme) — best unit: 145. vet bli-exactness-limits.bridge; 9 unit(s); Exact reflection fails on FAF's diagonal liar at interval resolution
  • bli-sources-refuted (shared claim of updateless-dt-programme) — best unit: 140. vet bli-sources-refuted.bridge; 9 unit(s); The post-hoc transfer of large prices
  • counterfactual-mugging-with-evidence (shared claim of updateless-dt-programme) — best unit: 204. vet counterfactual-mugging-with-evidence.bridge; 9 unit(s); UDT 1.1 behaves updatefully with respect to evidence exactly when the predictor's counterfactual reasoning is updateful with respect to it
  • cdt-defined-for-learner (shared claim of updateless-dt-programme) — best unit: 166. vet cdt-defined-for-learner.bridge; 9 unit(s); At a recorded point, interventional supposition on the act equals conditioning on it if and only if the act is independent of its graph-parents under the state, so the CDT/EDT…
  • two-lesions-propositions-checked (shared claim of updateless-dt-programme) — best unit: 196. vet two-lesions-propositions-checked.bridge; 9 unit(s); The propositions of the maintainer's two-lesions document
  • dp-dutch-book-done (shared claim of updateless-dt-programme) — best unit: 147. vet dp-dutch-book-done.bridge; 9 unit(s); The bet extension of a decision problem and who buys it
  • learning-versus-freezing (shared claim of updateless-dt-programme) — best unit: 179. vet learning-versus-freezing.bridge; 10 unit(s); In the finite shadow of a BLI prior
  • udt101-influence-calculus (shared claim of updateless-dt-programme) — best unit: 295. vet udt101-influence-calculus.bridge; 10 unit(s); The UDT 1.01 influence calculus of Posts 6 to 8 is formalized: Lemmas 1 and 2
  • harmony-as-bargaining (shared claim of updateless-dt-programme) — best unit: 304. vet harmony-as-bargaining.bridge; 10 unit(s); Harmony between heterogeneous decision points is a trembling-hand equilibrium of a bargaining game; the source note's Theorem 11.1 is false at two players and Proposition S replaces it
  • bli-picture-improvable (improvement claim on computational-updatelessness-direct, shared claim of updateless-dt-programme) — best unit: 161. vet updateful-when-appropriate.bridge; 46 unit(s); The BLI picture of computational updatelessness
  • bli-picture-improvable.residual (route under bli-picture-improvable) — best unit: 573. map-read updateful-when-appropriate.bridge; 9 unit(s); Some refinement of the literal plan not yet described reaches a version that holds for a computationally uncertain agent that observes its own future beliefs.
  • route-interval-resolution (route under bli-picture-improvable) — best unit: 161. vet updateful-when-appropriate.bridge; 40 unit(s); The exact constraints on large beliefs are relaxed to interval resolution
  • route-linkage-choice (route under bli-picture-improvable) — best unit: 161. vet updateful-when-appropriate.bridge; 40 unit(s); The linkage between the large sentences and the market's quotes
  • route-measure-valued-state (route under bli-picture-improvable) — best unit: 161. vet updateful-when-appropriate.bridge; 40 unit(s); The large sentences are replaced by a measure-valued state over a propositionally coherent inductor
  • route-scoped-faith (route under bli-picture-improvable) — best unit: 161. vet updateful-when-appropriate.bridge; 40 unit(s); Full-scope faith for B1 is dropped for a faith predicate scoped to the sentences the decision layer reads.
  • route-support-small-set (route under bli-picture-improvable) — best unit: 161. vet updateful-when-appropriate.bridge; 40 unit(s); The syntactic small set is replaced by the paper's support-based small set
  • li-bria-spec-improvable (improvement claim on indirect-route-learned-updatelessness, shared claim of updateless-dt-programme) — best unit: 116. vet policy-selection-over-li; 29 unit(s); The LI/BRIA acceptance statement with its updateless extension can be improved until it reflects a computable agent that learns policy-level behaviour well.
  • li-bria-spec-improvable.residual (route under li-bria-spec-improvable) — best unit: 386. map-read policy-selection-over-li; 7 unit(s); Some refinement of the literal plan not yet described reaches a version that holds for a computable agent that learns policy-level behaviour.
  • route-declared-commitment-source (route under li-bria-spec-improvable) — best unit: 116. vet policy-selection-over-li; 25 unit(s); The specification declares the candidate's commitment source
  • route-defined-recovery-domain (route under li-bria-spec-improvable) — best unit: 116. vet policy-selection-over-li; 25 unit(s); The updateless extension's recovery domain is defined by a property of the reward process stated without reference to the candidate agent.
  • route-one-shot-scope (route under li-bria-spec-improvable) — best unit: 116. vet policy-selection-over-li; 25 unit(s); The specification states whether one-shot verdicts are in scope
  • formal-core-improvable (improvement claim on decision-problems-formal-core, shared claim of updateless-dt-programme) — best unit: 143. vet formal-core-v3-trembling-hand.complete; 37 unit(s); The decision-problems formal core, trees over decision points with calibrated subjective states, can be improved until it reflects an embedded agent facing decision problems with predictors well.
  • formal-core-improvable.residual (route under formal-core-improvable) — best unit: 730. improve formal-core-improvable.residual; 15 unit(s); Some refinement of the literal plan not yet described reaches a version that holds for an embedded agent facing decision problems with predictors.
  • route-infinite-trees (route under formal-core-improvable) — best unit: 143. vet formal-core-v3-trembling-hand.complete; 31 unit(s); Finite trees are replaced by infinite trees with almost-surely finite runs
  • route-naturalized-agent (route under formal-core-improvable) — best unit: 143. vet formal-core-v3-trembling-hand.complete; 31 unit(s); The agent is embedded and naturalized
  • route-per-state-algebras (route under formal-core-improvable) — best unit: 143. vet formal-core-v3-trembling-hand.complete; 31 unit(s); The shared event algebra is replaced by per-state algebras with translations between them
  • route-shared-seed (route under formal-core-improvable) — best unit: 143. vet formal-core-v3-trembling-hand.complete; 31 unit(s); Independent redraws are replaced by shared-seed semantics or correspondences between decision points
  • route-state-dependent-exploration (route under formal-core-improvable) — best unit: 143. vet formal-core-v3-trembling-hand.complete; 31 unit(s); Uniform trembles are replaced by state-dependent exploration sequences
  • udt-tiling-improvable (improvement claim on udt-tiling-foundations, shared claim of updateless-dt-programme) — best unit: 418. re-examine udt-tiling-improvable; 38 unit(s); The Understanding Trust tiling framework can be improved until it reflects a self-modifying agent in time with computational uncertainty well.
  • route-computation-output-model (route under udt-tiling-improvable) — best unit: 425. vet self-coordination-udt10; 32 unit(s); Theorems 3 and 4 are read in a computation-output model in which Faith in Joint Argmax has content.
  • route-computational-uncertainty (route under udt-tiling-improvable) — best unit: 425. vet self-coordination-udt10; 32 unit(s); The tiling results are restated for computationally uncertain agents
  • route-explicit-causality (route under udt-tiling-improvable) — best unit: 425. vet self-coordination-udt10; 32 unit(s); Time and causality are explicit in the effective-policy operator
  • route-mean-fairness (route under udt-tiling-improvable) — best unit: 425. vet self-coordination-udt10; 32 unit(s); Policy Fairness is weakened to mean fairness where the two coincide on the paper's layer.
  • route-self-coordination-addressed (route under udt-tiling-improvable) — best unit: 425. vet self-coordination-udt10; 32 unit(s); UDT 1.0's self-coordination is addressed directly rather than assumed away
  • udt-tiling-improvable.residual (route under udt-tiling-improvable) — best unit: 753. map-read self-coordination-udt10; 10 unit(s); Some refinement of the literal plan not yet described reaches a version that holds for a self-modifying agent in time with computational uncertainty.
  • udt-rep-improvable (improvement claim on udt-representation-theorem, shared claim of updateless-dt-programme) — best unit: 324. re-examine udt-rep-improvable; 19 unit(s); The UDT representation-theorem framework, with decision-determination as its fairness axiom, can be improved until it reflects a system to which unified agency is correctly attributed well.
  • route-approximate-dd (route under udt-rep-improvable) — best unit: 352. improve route-approximate-dd; 15 unit(s); Exact decision-determination is replaced by its approximate form
  • route-minimality-premise (route under udt-rep-improvable) — best unit: 483. improve route-minimality-premise; 15 unit(s); A minimality premise is added under which the agency correspondence theorem is true
  • route-off-support-predictor (route under udt-rep-improvable) — best unit: 484. improve route-off-support-predictor; 15 unit(s); The predictor's access off the policy's support
  • udt-rep-improvable.residual (route under udt-rep-improvable) — best unit: 654. improve udt-rep-improvable.residual; 15 unit(s); Some refinement of the literal plan not yet described reaches a version that holds for a system to which unified agency is correctly attributed.
  • stance-and-level (frame, owner @abram): Where should the program work: at which level of computability, under which decision-theoretic stance, and does the computationally unbounded level represent the phenomena that matter?
  • level-super-turing (frame, owner @abram): Which work at the limit-computable level of AIXI and reflective oracles advances a formal theory of trust that guides AI safety engineering work?
  • factor-unbounded-then-relate (plan of level-super-turing) — best unit: 30. vet uea-floored-agent-theorem.bridge; 55 unit(s); Solving the embedded-agency and trust questions for unbounded computation first
  • reflective-oracles-correlated (plan of level-super-turing) — best unit: 207. vet crux-unbounded-represents-phenomena; 35 unit(s); A version of reflective oracles is constructed whose fixed-point operator is itself continuous and representable
  • vingean-uncertainty-is-computational (shared claim of level-super-turing) — best unit: 33. vet vingean-uncertainty-is-computational; 3 unit(s); Vingean uncertainty, trusting a decision procedure to perform well on general principles rather than by predicting it precisely, rules out computationally omniscient models, so it is uncertainty about computations.
  • uea-theorem-1 (shared claim of level-super-turing) — best unit: 1130. improve uea-theorem-1; 7 unit(s); There is a reflective oracle relative to which
  • uea-theorem-1.improvable (improvement claim on uea-theorem-1, shared claim of level-super-turing) — best unit: 30. vet uea-floored-agent-theorem.bridge; 21 unit(s); Cole Wyeth's revised Theorem 1 can be improved until it reflects a self-reflective agent that can modify itself and faces predictors of its policy well.
  • ro-independence-strain (route under uea-theorem-1.improvable) — best unit: 1130. improve uea-theorem-1; 8 unit(s); Replacing reflective oracles by fixed-point objects that allow correlations between program outputs yields a version of Theorem 1 under which an agent's choices can be modelled as correlated with a…
  • self-aixi-eps-exploration-route (route under uea-theorem-1.improvable) — best unit: 1130. improve uea-theorem-1; 8 unit(s); Trust-to-optimality for a self-predicting AIXI variant holds through ε-exploration rather than through the choice of reflective oracle
  • uea-current-node-floor-repair (route under uea-theorem-1.improvable) — best unit: 30. vet uea-floored-agent-theorem.bridge; 8 unit(s); Theorem 1 restated as a theorem about the current-node-floored agent over every reflective oracle holds in rOSI
  • uea-self-knowledge-vs-self-modification (route under uea-theorem-1.improvable) — best unit: 1130. improve uea-theorem-1; 8 unit(s); Theorem 1's self-knowledge premise is restated so that it distinguishes self-knowledge from environmental concerns
  • uea-theorem-1.residual (route under uea-theorem-1.improvable) — best unit: 960. improve uea-theorem-1.residual; 6 unit(s); Some refinement of Theorem 1 not yet described reflects a self-reflective agent that can modify itself and faces predictors of its policy.
  • uea-updateless-variant (route under uea-theorem-1.improvable) — best unit: 1130. improve uea-theorem-1; 8 unit(s); An updateless variant of Theorem 1
  • uea-floored-agent-theorem (refinement of uea-theorem-1) — best unit: 30. vet uea-floored-agent-theorem.bridge; 7 unit(s); In the finite shadow of Cole Wyeth's setting
  • uea-floored-agent-theorem.improvable (improvement claim on uea-floored-agent-theorem, shared claim of level-super-turing) — best unit: 30. vet uea-floored-agent-theorem.bridge; 11 unit(s); The floored-agent theorem can be improved until it reflects the reflective-oracle Solomonoff agent over an infinite horizon well.
  • uea-floor-log-form (route under uea-floored-agent-theorem.improvable) — best unit: 30. vet uea-floored-agent-theorem.bridge; 8 unit(s); The floored agent's value gap is at most O_h(3 + ln⁺(1/O_h)) at every floored fixed point
  • uea-floored-agent-theorem.residual (route under uea-floored-agent-theorem.improvable) — best unit: 30. vet uea-floored-agent-theorem.bridge; 6 unit(s); Some refinement of the floored-agent theorem not yet described reflects the reflective-oracle Solomonoff agent over an infinite horizon.
  • uea-rosi-lift (route under uea-floored-agent-theorem.improvable) — best unit: 30. vet uea-floored-agent-theorem.bridge; 8 unit(s); Theorems B and C of the finite sequential self-game hold in the reflective-oracle Solomonoff setting with an infinite discounted horizon
  • uea-t1-key-lemma-refuted-finite (shared claim of level-super-turing) — best unit: 99. vet uea-t1-key-lemma-refuted-finite.bridge; 7 unit(s); In the finite sequential self-game
  • uea-post-usc-backwards (shared claim of level-super-turing) — best unit: 89. vet uea-post-usc-backwards.bridge; 7 unit(s); The live post's argument that the modified best-response map stays upper semicontinuous fails: a strict reset inequality makes the reset region open
  • uea-sink-or-swim-trap (shared claim of level-super-turing) — best unit: 119. vet uea-sink-or-swim-trap.bridge; 7 unit(s); In the sink-or-swim model with a fixed non-dogmatic prior
  • uea-printed-2025-refuted (shared claim of level-super-turing) — best unit: 87. vet uea-printed-2025-refuted.bridge; 7 unit(s); The 2025 printing of Theorem 1
  • uea-horizon-quantifiers (shared claim of level-super-turing) — best unit: 115. vet uea-horizon-quantifiers.bridge; 7 unit(s); In the finite shadow, for every ε and horizon T there is a δ that makes the self-trusting agent ε-optimal, but without discounting (γ = 1) no single δ serves…
  • udt10-believes-udt11 (shared claim of level-super-turing) — best unit: 172. vet udt10-believes-udt11; 22 unit(s); If a UDT1.0 agent assigns probability at least 1 − δ to its policy being UDT1.1
  • udt-trust-existence-and-tightness (shared claim of level-super-turing) — best unit: 52. vet udt-trust-existence-and-tightness.bridge; 7 unit(s); In the updateless self-game, extension fixed points exist for every game; with the trust bound at every situation the supremum of the gap is U* − δ/(1 − δ) on…
  • ro-generalized-kakutani-refuted (shared claim of level-super-turing) — best unit: 123. vet ro-generalized-kakutani-refuted.bridge; 7 unit(s); The reflective-oracle project's generalized Kakutani claim
  • ro-ftc-finite-core-from-brouwer (shared claim of level-super-turing) — best unit: 86. vet ro-ftc-finite-core-from-brouwer.bridge; 7 unit(s); The finite core of Fallenstein
  • uea-improbable-actions (shared claim of level-super-turing) — best unit: none ranked; 2 unit(s); The self-reflective agent's reasoning about low-probability actions causes no problem of the 5-and-10 kind once it is mostly confident of its source code
  • want-uea-review (want of @abram) — best unit: 1851. spell-out want-uea-review; 2 unit(s); Cole Wyeth's revised Theorem 1 (rOSI) is assessed in a written account that places it against tiling
  • want-methodology-note (want of @abram) — best unit: 1797. do factor-methodology-note; 3 unit(s); The case for and against working at the computationally unbounded level
  • level-above-quantifiable (frame, owner @abram): Which work at the level of probability frames and Jeffrey–Bolker models advances a formal theory of trust that guides AI safety engineering work?
  • aq-frame-level-then-translate (plan of level-above-quantifiable) — best unit: 64. vet ddb-finite-frames-formalized.bridge; 39 unit(s); Working trust principles out on probability frames and in the Jeffrey–Bolker format
  • ddb-value-iff-total-trust (shared claim of level-above-quantifiable) — best unit: 64. vet ddb-finite-frames-formalized.bridge; 11 unit(s); On finite probability frames, a novice values letting an expert decide (Value) exactly when it totally trusts the expert's estimates (Total Trust).
  • ddb-value-iff-total-trust.improvable (improvement claim on ddb-value-iff-total-trust, shared claim of level-above-quantifiable) — best unit: 64. vet ddb-finite-frames-formalized.bridge; 17 unit(s); Deference Done Better's characterization of rational deference can be improved until it reflects rational deference between embedded agents well.
  • ddb-value-iff-total-trust.residual (route under ddb-value-iff-total-trust.improvable) — best unit: 423. map-read ddb-finite-frames-formalized.bridge; 2 unit(s); Some generalization of Deference Done Better's characterization not yet described reflects rational deference between embedded agents.
  • frames-beyond-finite (route under ddb-value-iff-total-trust.improvable) — best unit: 64. vet ddb-finite-frames-formalized.bridge; 12 unit(s); On countable or uncountable frames an appropriately weakened Value is equivalent to an appropriately weakened Total Trust.
  • frames-with-self-reference (route under ddb-value-iff-total-trust.improvable) — best unit: 64. vet ddb-finite-frames-formalized.bridge; 12 unit(s); A representation of beliefs about beliefs that admits Gödelian self-reference supports a weakened form of Deference Done Better's results.
  • value-jeffrey-bolker-format (route under ddb-value-iff-total-trust.improvable) — best unit: 64. vet ddb-finite-frames-formalized.bridge; 12 unit(s); Value restated in a Jeffrey–Bolker format
  • value-symmetric-comparator (route under ddb-value-iff-total-trust.improvable) — best unit: 64. vet ddb-finite-frames-formalized.bridge; 12 unit(s); Value with the novice's choice also allowed to depend on the world keeps a characterization by a trust principle on finite frames.
  • infinite-frames-sentence-check (shared claim of level-above-quantifiable) — best unit: 780. do infinite-frames-sentence-check; 4 unit(s); The sentence of li-deference.md that the Value–Total Trust result fails for infinite frames in both directions holds for the menu class the maintainer means
  • tt-finite-frames-settled (shared claim of level-above-quantifiable) — best unit: 50. vet tt-finite-frames-settled.bridge; 7 unit(s); On finite frames with one prior per agent
  • mm-theorem-3-4-refuted (shared claim of level-above-quantifiable) — best unit: 96. vet mm-theorem-3-4-refuted.bridge; 7 unit(s); Managing Misalignment's Theorem 3.4 is false in both directions as printed
  • level-poly-time (frame, owner @abram; not opened): Which whole-mind models of rationality at polynomial time or below would advance a formal theory of trust?
  • crux-unbounded-represents-phenomena (shared claim of stance-and-level) — best unit: 207. vet crux-unbounded-represents-phenomena; 13 unit(s); The phenomena that matter for trust between humans and AI systems
  • super-turing-guides-engineering (shared claim of stance-and-level) — best unit: 207. vet crux-unbounded-represents-phenomena; 21 unit(s); A formal theory of trust developed at the limit-computable level guides AI safety engineering work.
  • li-guides-engineering (shared claim of stance-and-level) — best unit: 233. vet li-results-reach-selected-systems; 24 unit(s); A formal theory of trust developed at the logical-induction level guides AI safety engineering work.
  • above-quantifiable-guides-engineering (shared claim of stance-and-level) — best unit: 283. vet aq-guides-via-translation; 10 unit(s); A formal theory of trust developed at the level of probability frames and Jeffrey–Bolker models guides AI safety engineering work.
  • value-uncertainty-is-computational (shared claim of stance-and-level) — best unit: 1104. do value-uncertainty-is-computational; 2 unit(s); Human uncertainty about human values is to a substantial degree computational uncertainty
  • li-updatelessness-tractable (shared claim of stance-and-level) — best unit: 200. vet li-updatelessness-tractable; 4 unit(s); A computationally uncertain UDT exists over a logical-induction-like belief process
  • updateless-engineering-value (shared claim of stance-and-level) — best unit: 238. vet updateless-engineering-value; 6 unit(s); The updateless side is high value to AI safety engineering work even where the theory of computational updatelessness is easy: the engineering decisions updateless reasoning changes are large enough that…
  • abstraction-arc (frame, owner @abram): Which work on abstraction, logical induction and agency gives the program a representation theorem for logical inductors, a decision theory set in that representation, and a notion of scale-free agency that covers cooperative equilibria?
  • want-intentional-stance-math (want of @abram) — best unit: 1843. spell-out want-intentional-stance-math; 1 unit(s); A mathematical characterization states when it is rational to regard observed phenomena as agentic
  • stage4-condensation-dynamics (frame, owner @abram): Which notions of function and isomorphism, which role for the observables, and which dynamics of theories carry the same concept between agents and across ontological shifts?
  • condensation-correspondence-theorem (shared claim of stage4-condensation-dynamics) — best unit: 71. vet condensation-correspondence-theorem.bridge; 11 unit(s); Two latent-variable models that both perfectly condense one random-variable model stand in correspondence: each latent of either model is almost everywhere a function of the other model's latents above the…
  • condensation-correspondence-theorem.improvable (improvement claim on condensation-correspondence-theorem, shared claim of stage4-condensation-dynamics) — best unit: 71. vet condensation-correspondence-theorem.bridge; 15 unit(s); Eisenstat's correspondence theorem for condensation can be improved until it reflects concepts shared between reasoning agents and kept across changes of theory well.
  • condensation-correspondence-theorem.residual (route under condensation-correspondence-theorem.improvable) — best unit: 71. vet condensation-correspondence-theorem.bridge; 7 unit(s); Some refinement of Eisenstat's correspondence theorem not described by the named routes reflects concepts shared between reasoning agents and kept across changes of theory.
  • route-condensation-algorithmic (route under condensation-correspondence-theorem.improvable) — best unit: 71. vet condensation-correspondence-theorem.bridge; 7 unit(s); An algorithmic-information version of condensation
  • route-condensation-computational-cost (route under condensation-correspondence-theorem.improvable) — best unit: 71. vet condensation-correspondence-theorem.bridge; 7 unit(s); A version of condensation that charges for computation as well as retrieval
  • route-condensation-dynamic (route under condensation-correspondence-theorem.improvable) — best unit: 71. vet condensation-correspondence-theorem.bridge; 7 unit(s); A time-indexed version of condensation
  • route-condensation-givens-morphism (route under condensation-correspondence-theorem.improvable) — best unit: 71. vet condensation-correspondence-theorem.bridge; 7 unit(s); A variant of the correspondence theorem holds for two latent models whose given variables are related by a morphism rather than shared
  • route-condensation-relational-morphisms (route under condensation-correspondence-theorem.improvable) — best unit: 71. vet condensation-correspondence-theorem.bridge; 7 unit(s); A variant of the correspondence theorem holds for a notion of morphism more general than condensation's single-source maps.
  • route-condensation-storage-retrieval (route under condensation-correspondence-theorem.improvable) — best unit: 71. vet condensation-correspondence-theorem.bridge; 7 unit(s); A version of condensation that trades storage cost against retrieval cost
  • route-condensation-support-sensitive (route under condensation-correspondence-theorem.improvable) — best unit: 71. vet condensation-correspondence-theorem.bridge; 7 unit(s); A support-sensitive notion of equivalence
  • condensation-equivalence-ae (shared claim of stage4-condensation-dynamics) — best unit: 92. vet condensation-equivalence-ae.bridge; 7 unit(s); In Eisenstat's category of random-variable models an equivalence is an isomorphism almost everywhere
  • condensation-morphism-notion (plan of stage4-condensation-dynamics) — best unit: 569. re-examine condensation-morphism-notion; 6 unit(s); Some notion of morphism between random-variable models beyond Eisenstat's single-source index maps
  • observables-carry-correspondence (shared claim of stage4-condensation-dynamics) — best unit: 1640. do own-posts-as-todos; 3 unit(s); The given variables two latent-variable models share do part of the work that makes their correspondence meaningful
  • interpreting-a-belief-haver (plan of stage4-condensation-dynamics) — best unit: 607. re-examine interpreting-a-belief-haver; 7 unit(s); A representation of one agent interpreting another as a belief-haver
  • dynamics-of-theories (plan of stage4-condensation-dynamics) — best unit: 594. re-examine dynamics-of-theories; 5 unit(s); A time-indexed version of condensation states how concepts persist and deform as a theory changes and what survives an ontological shift.
  • superconditioning-calibration.improvable (improvement claim on superconditioning-calibration, shared claim of stage4-condensation-dynamics) — best unit: 681. re-examine superconditioning-calibration.improvable; 9 unit(s); The superconditioning account of calibration across ontologies can be improved until it reflects ontological shifts in real reasoners well.
  • route-superconditioning-approximate-common-information (route under superconditioning-calibration.improvable) — best unit: 826. improve route-superconditioning-approximate-common-information; 4 unit(s); Replacing the exact homomorphisms of the common-information structure by approximate ones yields a calibration theory for imperfect ontology translation.
  • route-superconditioning-cross-ontology-utility (route under superconditioning-calibration.improvable) — best unit: 886. improve route-superconditioning-cross-ontology-utility; 4 unit(s); A rule for transforming utilities across ontologies
  • route-superconditioning-logical-uncertainty (route under superconditioning-calibration.improvable) — best unit: 1032. improve route-superconditioning-logical-uncertainty; 4 unit(s); Extending the framework so that the world-model itself is logically uncertain
  • route-superconditioning-proof-gaps-repaired (route under superconditioning-calibration.improvable) — best unit: 689. improve route-superconditioning-proof-gaps-repaired; 4 unit(s); Repairing the document's catalogued proof gaps yields a version whose theorems all hold as stated.
  • superconditioning-calibration.residual (route under superconditioning-calibration.improvable) — best unit: 950. improve superconditioning-calibration.residual; 4 unit(s); Some refinement of the superconditioning account not described by the named routes reflects ontological shifts in real reasoners.
  • implementation-relations (plan of stage4-condensation-dynamics) — best unit: 296. re-examine implementation-relations; 12 unit(s); A notion of one structure implementing another represents objects that move between variables without letting arbitrary computation make every structure implement every other.
  • abstraction-theories-integration (plan of stage4-condensation-dynamics) — best unit: 448. re-examine abstraction-theories-integration; 8 unit(s); Petersen's theory of abstraction, Wentworth's natural abstraction and natural latents, and Eisenstat's condensation are related precisely: Petersen's theory to Wentworth's, natural latents to condensation, and Petersen's theory to condensation, each…
  • condensation-scores-explained (plan of stage4-condensation-dynamics) — best unit: 183. vet two-scores-as-explanation-length; 7 unit(s); Each of condensation's three scores is interpreted as an expected cost under a stated convention: the simple and conditioned scores as explanation lengths
  • condensation-algorithmic-version (plan of stage4-condensation-dynamics) — best unit: 582. re-examine condensation-algorithmic-version; 4 unit(s); An algorithmic version of condensation
  • notebooks-that-call-each-other (shared claim of stage4-condensation-dynamics) — best unit: 1182. improve notebooks-that-call-each-other; 2 unit(s); A variation of condensation in which notebooks may contain instructions for retrieving other notebooks has a correspondence theorem.
  • permissive-morphisms-needed (shared claim of stage4-condensation-dynamics) — best unit: 177. vet permissive-morphisms-needed; 4 unit(s); Condensation's single-source morphisms suffice to carry the same concept between agents.
  • stage5-li-representation-theorem (frame, owner @abram): Which abstraction of a logically uncertain belief state, and which simplicity argument, make it natural to interpret an observed system as a logical inductor?
  • li-representation-theorem (plan of stage5-li-representation-theorem) — best unit: 472. vet li-representation-theorem.complete; 6 unit(s); Under stated conditions on an observed process
  • li-representation-theorem.improvable (improvement claim on li-representation-theorem, shared claim of stage5-li-representation-theorem) — best unit: 472. vet li-representation-theorem.complete; 12 unit(s); The representation theorem for logical inductors can be improved until it reflects systems deployed as AI agents well.
  • li-representation-theorem.residual (route under li-representation-theorem.improvable) — best unit: 472. vet li-representation-theorem.complete; 7 unit(s); Some refinement of the representation theorem not described by the named routes reflects systems deployed as AI agents.
  • route-li-rep-agent-structure-analogy (route under li-representation-theorem.improvable) — best unit: 472. vet li-representation-theorem.complete; 7 unit(s); An agent-structure theorem, stating that a system optimizing at some level of abstraction is approximately a logical inductor there, reflects deployed AI systems.
  • route-li-rep-generalized-criterion (route under li-representation-theorem.improvable) — best unit: 472. vet li-representation-theorem.complete; 7 unit(s); Stating the theorem over a generalized logical-induction criterion
  • route-li-rep-via-endorsement (route under li-representation-theorem.improvable) — best unit: 472. vet li-representation-theorem.complete; 7 unit(s); Routing the theorem through endorsement
  • route-li-rep-via-vingean (route under li-representation-theorem.improvable) — best unit: 472. vet li-representation-theorem.complete; 7 unit(s); Starting from Vingean agency, predicting a stronger agent's outcome without its moves, gives a representation theorem whose hypotheses deployed systems meet.
  • li-compatible-abstraction-theory (plan of stage5-li-representation-theorem) — best unit: 292. re-examine li-compatible-abstraction-theory; 9 unit(s); A theory of abstraction exists whose abstracted object is a logically uncertain belief state
  • inferential-semantics-for-logical-udt (plan of stage5-li-representation-theorem) — best unit: 635. re-examine inferential-semantics-for-logical-udt; 7 unit(s); A tiling theorem for logically uncertain updateless decision theory establishes trust in a future self through three parts: a market-derivative algebra of logical uncertainty giving an inferential semantics
  • agency-correspondence-refuted (shared claim of stage5-li-representation-theorem) — best unit: 133. vet agency-correspondence-refuted.complete; 8 unit(s); Under determination of actions by policy and observation and recoverability of the policy from behaviour alone
  • dd-is-conditional-independence (shared claim of stage5-li-representation-theorem) — best unit: 107. vet dd-is-conditional-independence.bridge; 7 unit(s); Decision-determination is the conditional independence of the environment from decisions given the policy latent
  • vingean-agency-via-abstraction (plan of stage5-li-representation-theorem) — best unit: 299. re-examine vingean-agency-via-abstraction; 9 unit(s); The epistemic position of predicting that a more skilled agent will win without predicting its moves is characterized by a theory of abstraction.
  • vingean-as-blackwell-dominance.improvable (improvement claim on vingean-as-blackwell-dominance, shared claim of stage5-li-representation-theorem) — best unit: 301. vet vingean-as-blackwell-dominance; 9 unit(s); The Blackwell-dominance account of Vingean agency can be improved until it reflects the epistemic position of humans overseeing stronger AI systems well.
  • route-vingean-condensation-of-outcome (route under vingean-as-blackwell-dominance.improvable) — best unit: 301. vet vingean-as-blackwell-dominance; 6 unit(s); Vingean agency is possessing a condensation of the agent's outcome variable but not of its policy latent.
  • route-vingean-menu-robust (route under vingean-as-blackwell-dominance.improvable) — best unit: 301. vet vingean-as-blackwell-dominance; 6 unit(s); Requiring the value-of-information dominance across a family of menus excludes the thermostat and keeps chess-like Vingean agents.
  • vingean-as-blackwell-dominance.residual (route under vingean-as-blackwell-dominance.improvable) — best unit: 301. vet vingean-as-blackwell-dominance; 6 unit(s); Some refinement of the Blackwell-dominance account not described by the named routes reflects the epistemic position of humans overseeing stronger AI systems.
  • intentional-stance-characterized (shared claim of stage5-li-representation-theorem) — best unit: none ranked; 2 unit(s); Conditions on observed phenomena under which it is rational to regard them as agentic are stated so that both the representation-theorem route and the Vingean route are instances
  • agency-via-endorsement-route (shared claim of stage5-li-representation-theorem) — best unit: 324. re-examine udt-rep-improvable; 20 unit(s); Updateless decision theory is what being control-endorsed by one's own prior means
  • cartesian-frames-of-li (shared claim of stage5-li-representation-theorem) — best unit: 1157. improve cartesian-frames-of-li; 2 unit(s); A Cartesian-frames-like decomposition of a logical-induction world into many interacting sub-markets exists.
  • unify-abstraction-with-tiling (plan of stage5-li-representation-theorem) — best unit: 224. vet unify-abstraction-with-tiling; 4 unit(s); The unified theory of abstraction and agents is joined with the tiling line into one theory.
  • li-abstraction-tractable (shared claim of stage5-li-representation-theorem) — best unit: 174. vet li-abstraction-tractable; 17 unit(s); A theory of abstraction compatible with logical induction exists without changing the logical induction criterion.
  • vingean-or-representation-start (shared claim of stage5-li-representation-theorem) — best unit: 156. vet vingean-or-representation-start; 4 unit(s); Vingean agency and the representation theorem for logical inductors are one problem under two descriptions
  • stage6-dt-and-scale-free-agency (frame, owner @abram): Which decision theory, set in the Stage 5 object, and which notion of scale-free agency characterize approximate cooperative equilibria as agents, with advice for steering toward them?
  • decision-theory-in-stage5-setting (plan of stage6-dt-and-scale-free-agency) — best unit: 186. vet stage5-right-setting-for-dt; 7 unit(s); The decision-problems formal core is restated with the Stage 5 object as the decision procedure
  • scale-free-agency (plan of stage6-dt-and-scale-free-agency) — best unit: 1011. vet scale-free-agency.complete; 5 unit(s); A notion of agency that applies at every scale of description is derived from a theory of the abstraction of agency
  • cooperative-equilibria-as-agents (frame, owner @abram): Which characterization of approximate cooperative equilibria among learning agents as agents, with enforcement against an AI leaving the coalition, yields advice for steering toward them?
  • aggregate-logical-inductor (shared claim of cooperative-equilibria-as-agents) — best unit: 121. vet aggregate-logical-inductor; 4 unit(s); Members entering the logical-induction paper's fixed-point market as Kelly-demand traders with every budget
  • aggregate-logical-inductor.improvable (improvement claim on aggregate-logical-inductor, shared claim of cooperative-equilibria-as-agents) — best unit: 121. vet aggregate-logical-inductor; 11 unit(s); The coalition-as-market picture of the aggregate logical inductor can be improved until it reflects a human–AI coalition with learning members
  • aggregate-logical-inductor.residual (route under aggregate-logical-inductor.improvable) — best unit: 121. vet aggregate-logical-inductor; 5 unit(s); Some refinement of the coalition-as-market picture not described by the named routes reflects a human–AI coalition with learning members
  • route-coalition-cap-maintained (route under aggregate-logical-inductor.improvable) — best unit: 121. vet aggregate-logical-inductor; 5 unit(s); A maintained cap on the AI's weight
  • route-coalition-decision-layer (route under aggregate-logical-inductor.improvable) — best unit: 121. vet aggregate-logical-inductor; 5 unit(s); A decision layer in which each member's own expectation drives the policy with the aggregate as bookmaker keeps Pareto optimality where the aggregate's expectation loses it.
  • route-coalition-disclosure-regime (route under aggregate-logical-inductor.improvable) — best unit: 121. vet aggregate-logical-inductor; 5 unit(s); Fixing a disclosure regime (prices only, prices and wealths, or the full record) makes the secrecy and detection results determinate.
  • route-coalition-dynamic-weight-law (route under aggregate-logical-inductor.improvable) — best unit: 121. vet aggregate-logical-inductor; 5 unit(s); A law for how bargaining weight moves between learning members replaces Critch's fixed-prior recursion and keeps the weight-path result.
  • route-coalition-literal-exit (route under aggregate-logical-inductor.improvable) — best unit: 121. vet aggregate-logical-inductor; 5 unit(s); Modelling the participation constraint with literal exit
  • self-enforcing-coalition-conditions (plan of cooperative-equilibria-as-agents) — best unit: 638. re-examine self-enforcing-coalition-conditions; 7 unit(s); A coalition of logical-inductor or BRIA agents communicating through a shared market is a self-enforcing equilibrium under stated conditions
  • deference-protects-from-outprediction-not-deception (shared claim of cooperative-equilibria-as-agents) — best unit: 36. vet deference-protects-from-outprediction-not-deception; 4 unit(s); A deferring member's weight ratio is a martingale under the stronger member's beliefs whatever that member announces
  • acts-of-war-as-re-bargains (shared claim of cooperative-equilibria-as-agents) — best unit: 93. vet acts-of-war-as-re-bargains; 7 unit(s); An act of war in the coalition is a unilateral
  • secrecy-not-forbidden-by-criterion (shared claim of cooperative-equilibria-as-agents) — best unit: 117. vet secrecy-not-forbidden-by-criterion; 4 unit(s); Over the announced record the logical induction criterion forbids no secrecy; over the union of what members know it fails for any permanently unannounced independent fact; withholding policies visible in…
  • hundred-percent-split-is-li-deference (shared claim of cooperative-equilibria-as-agents) — best unit: 282. vet hundred-percent-split-is-li-deference; 5 unit(s); The coalition with all bargaining weight on the humans is the li-deference program: faithful acceleration on the settled fragment and legitimacy on the never-settled one.
  • coalition-detects-ai-values (shared claim of cooperative-equilibria-as-agents) — best unit: 239. vet coalition-detects-ai-values; 5 unit(s); The coalition's shared market can tell that an AI member has values of its own to suppress.
  • coalition-object-static-or-dynamic (shared claim of cooperative-equilibria-as-agents) — best unit: 263. vet coalition-object-static-or-dynamic; 4 unit(s); Results proved for Critch's static recursion apply to the coalition of the roadmap's Stage 6
  • informational-coup-is-act-of-war (shared claim of cooperative-equilibria-as-agents) — best unit: 100. vet informational-coup-is-act-of-war; 4 unit(s); A dramatic cash-in of long-withheld information is an act of war in the coalition picture
  • ledger-logic-market-logic-synthesis (frame, owner @abram): Which results of a ledger formulation and of Market Logic hold as one body of mathematics, and where do the two frameworks part?
  • market-maker-is-li-maker-over-goods.improvable (improvement claim on market-maker-is-li-maker-over-goods, shared claim of ledger-logic-market-logic-synthesis) — best unit: 711. re-examine market-maker-is-li-maker-over-goods.improvable; 9 unit(s); Market Logic's market over goods can be improved until it reflects belief and logic dynamics with a proof theory well
  • market-maker-is-li-maker-over-goods.residual (route under market-maker-is-li-maker-over-goods.improvable) — best unit: 953. improve market-maker-is-li-maker-over-goods.residual; 4 unit(s); Some refinement of Market Logic's market not described by the named routes reflects belief and logic dynamics with a proof theory.
  • route-market-classical-goods (route under market-maker-is-li-maker-over-goods.improvable) — best unit: 805. improve route-market-classical-goods; 4 unit(s); A condition on the goods and traders of Market Logic I's market-maker
  • route-market-computable-fixed-point (route under market-maker-is-li-maker-over-goods.improvable) — best unit: 1022. vet market-maker-is-li-maker-over-goods; 4 unit(s); A computable market-maker fixed point exists with goods and traders added daily.
  • route-market-proof-theory (route under market-maker-is-li-maker-over-goods.improvable) — best unit: 954. improve route-market-proof-theory; 4 unit(s); A proof theory for the derivative logic exists and is the logic of the post 'Infinitesimally False'.
  • route-market-substitution-from-trades (route under market-maker-is-li-maker-over-goods.improvable) — best unit: 955. improve route-market-substitution-from-trades; 4 unit(s); Substitution rules can be read off the market's trade patterns.
  • never-settled-standard (shared claim of ledger-logic-market-logic-synthesis) — best unit: none ranked; 2 unit(s); The standard governing prices on never-settled questions comes from a legitimacy-modified version of the logical-induction trust principles (Total Trust, the tower or Reflection)
  • stage5-right-setting-for-dt (shared claim of abstraction-arc) — best unit: 186. vet stage5-right-setting-for-dt; 4 unit(s); Decision theory formulated over the Stage 5 object
  • mastery-gates-stage5 (shared claim of abstraction-arc) — best unit: 85. vet mastery-gates-stage5; 4 unit(s); Formal work on the Stage 5 representation theorem proceeds in parallel with the mastery program rather than after it.
  • enabling-work (frame, owner @anson, @abram; holds 15): Which repository infrastructure makes the trust research vettable and publishable?
  • enabling-vetting-and-prioritization-system (shared claim of enabling-work) — best unit: 605. vet ranking-matches-stated-priorities; 16 unit(s); The argument map under graph/ files the research agenda from deep goals down to concrete tasks
  • enabling-machine-checked-corpus (shared claim of enabling-work) — best unit: 873. vet cleanroom-results-in-public-ci; 13 unit(s); The research corpus's claims are formalized in Lean 4 over FAF and adversarially audited in a run independent of the first
  • enabling-publication-pipeline (shared claim of enabling-work) — best unit: 271. vet enabling-publication-pipeline.complete; 6 unit(s); Each candidate for release to this repository is listed with its state
  • crux-formalize-before-write-up (shared claim of enabling-work) — best unit: 141. vet crux-formalize-before-write-up; 4 unit(s); Finishing the clean-room repair round and the note corrections before drafting the li-deference write-up gives a released write-up that needs fewer corrections after release
  • crux-vetting-or-generation (shared claim of enabling-work) — best unit: 139. vet crux-vetting-or-generation; 4 unit(s); Over the next researcher-hours of the agenda
• instrumental-convergence (risk story of top) — best unit: 61. vet ps-cor614-orbit-count.complete; 35 unit(s); Several instrumental values are convergent: their attainment would increase the chances of the agent's goal being realized for a wide range of final goals and situations
• corrigibility-anti-natural (risk story of top) — best unit: 464. vet corrigibility-anti-natural.complete; 9 unit(s); Almost all rational agents are instrumentally motivated to preserve their preferences
• dishonest-inner-optimizers (risk story of top) — best unit: 298. vet misreport-induces-action; 10 unit(s); Malign hypotheses in a value learner's broad prior gain weight by predicting or trading well
• legit-objective-li-reading (route under legit-weighted-objective-improvable) — best unit: 77. vet void-branch-deterrence-floor.bridge; 22 unit(s); The five objectives stated at the logical-induction register
• legit-relax-anchor (route under legit-trust-improvable) — best unit: 97. vet legit-relax-anchor; 9 unit(s); With the legitimacy event anchored in the default overseer process or an idealized judge
• legit-relax-finite-frames (route under legit-trust-improvable) — best unit: 959. vet legit-ddb-transfer.bridge; 7 unit(s); The L-conditioned Total Trust equivalences hold on the general object with regular conditional probabilities
• legit-rule-learning-model (route under legit-weighted-objective-improvable) — best unit: 77. vet void-branch-deterrence-floor.bridge; 22 unit(s); A model of corrigibility rules learned from legitimate-branch feedback measures the selection bias the ex-ante conditioning result predicts
• legit-trust-improvable (improvement claim on legit-ddb-transfer) — best unit: 97. vet legit-relax-anchor; 19 unit(s); The finite-frame L-conditioned Total Trust theory can be improved until it reflects the teaching channel between a human process and an AI well.
• legit-trust-improvable.residual (route under legit-trust-improvable) — best unit: 1146. improve legit-trust-improvable.residual; 1 unit(s); The finite theory improves toward the real channel by a route none of the named relaxations describes.
• legit-v1-improvable (improvement claim on legit-endpoint-preservation) — best unit: 37. vet route-v1-timely-fragment-certificate; 17 unit(s); The endpoint-preservation model can be improved until it reflects the legitimacy of a real advisor-human channel well.
• legit-v1-improvable.residual (route under legit-v1-improvable) — best unit: 1177. improve legit-v1-improvable.residual; 1 unit(s); The endpoint-preservation model improves toward real advisor-human legitimacy by a route none of the named ones describes.
• legit-weighted-objective-improvable (improvement claim on legit-objective-proposals) — best unit: 77. vet void-branch-deterrence-floor.bridge; 34 unit(s); The legitimacy-weighted decision objectives can be improved until they reflect how legitimacy considerations enter an AI's choices well.
• legit-weighted-objective-improvable.residual (route under legit-weighted-objective-improvable) — best unit: 1176. improve legit-weighted-objective-improvable.residual; 1 unit(s); The legitimacy-weighted objectives improve toward how legitimacy enters an AI's choices by a route none of the named ones describes.
• reasons-improvable (improvement claim on reasons-statics-public) — best unit: 129. vet route-reasons-relative-prices; 13 unit(s); The finite reasons-and-warrant statics can be improved until they reflect how human reasons function in legitimate deliberation well.
• reasons-improvable.residual (route under reasons-improvable) — best unit: 1141. improve reasons-improvable.residual; 1 unit(s); The reasons statics improve toward how human reasons function by a route none of the named ones describes.
• route-piecemeal-design (route under legit-weighted-objective-improvable) — best unit: 246. vet route-piecemeal-design; 6 unit(s); A system that is consequentialist by default and corrigible on legitimate modification attempts
• route-reasons-content-theory (route under reasons-improvable) — best unit: 1098. improve route-reasons-content-theory; 4 unit(s); The statics gain a theory of content
• route-reasons-items-54-58 (route under reasons-improvable) — best unit: 1183. vet reasons-statics-public; 4 unit(s); The reason-state blockers of PRIORITIES items 54 to 58 are closed: defeater-uptake completeness
• route-reasons-relative-prices (route under reasons-improvable) — best unit: 129. vet route-reasons-relative-prices; 6 unit(s); A scorekeeper's perspective is modelled as a market with reader-specific wealth
• route-t3-second-decision (route under legit-weighted-objective-improvable) — best unit: 77. vet void-branch-deterrence-floor.bridge; 22 unit(s); In a sequential problem with a second legitimacy decision
• route-v1-defect-as-thin-correct (route under legit-v1-improvable) — best unit: 44. vet route-v1-defect-as-thin-correct; 5 unit(s); The influence defect tending to zero is the LI-register counterpart of the thin level's correct property with the sealed sibling as the baseline
• route-v1-second-channel (route under legit-v1-improvable) — best unit: none ranked; 1 unit(s); A second channel, an influence map in the record, gives detection of endpoint movement and softens the acceleration-only thesis to admit legitimate uplift on never-settled questions.
• route-v1-timely-fragment-certificate (route under legit-v1-improvable) — best unit: 37. vet route-v1-timely-fragment-certificate; 6 unit(s); Endpoint preservation restricted to the timely fragment

Judgment and structure units (unranked)

In order: the owners' priority words, gestalts that set a goal's promise, estimates (each points at the read that supersedes it), then classification, quick admission, reconciliation, approval and structural re-examination. informs k goals counts the goals whose inference reaches the unit's node.

• priority — 4 wants of @anson await a priority word (high, medium, low or none; unset reads as medium on a want you filed, as nothing on one quoted from your words), first want-address-anti-naturality, want-formal-model-of-dynamics, want-sigma-algebra-changes; --as @anson lists each with its source
• priority — 9 wants of @abram await a priority word (high, medium, low or none; unset reads as medium on a want you filed, as nothing on one quoted from your words), first want-figure-out-legitimacy, want-legitimacy-meets-syntax, want-basin-of-attraction-formalized; --as @abram lists each with its source
• gestalt public-legitimacy-theory — informs 1 goal; the derived number multiplies 3 unread guesses; your holistic number for public-legitimacy-theory (about two minutes) sets its promise — under public-legitimacy-theory
• gestalt wiki-corpus-maintenance — informs 1 goal; the derived number multiplies 7 unread guesses; your holistic number for wiki-corpus-maintenance (about two minutes) sets its promise — under wiki-corpus-maintenance
• gestalt deck-collapse-statement — informs 7 goals; the derived number multiplies 5 unread guesses; your holistic number for deck-collapse-statement (about two minutes) sets its promise — under deck-collapse-statement
• gestalt counterfactuals-for-embedded-agents — informs 1 goal; the derived number multiplies 3 unread guesses; your holistic number for counterfactuals-for-embedded-agents (about two minutes) sets its promise — under counterfactuals-for-embedded-agents
• gestalt li-deference-collapse-scoped — informs 8 goals; the derived number multiplies 7 unread guesses; your holistic number for li-deference-collapse-scoped (about two minutes) sets its promise — under li-deference-collapse-scoped
• gestalt uea-floored-agent-theorem — informs 8 goals; the derived number multiplies 3 unread guesses; your holistic number for uea-floored-agent-theorem (about two minutes) sets its promise — under uea-floored-agent-theorem
• gestalt authority-typed-corrigibility — informs 6 goals; the derived number multiplies 4 unread guesses; your holistic number for authority-typed-corrigibility (about two minutes) sets its promise — under authority-typed-corrigibility
• gestalt inferential-semantics-for-logical-udt — informs 1 goal; the derived number multiplies 4 unread guesses; your holistic number for inferential-semantics-for-logical-udt (about two minutes) sets its promise — under inferential-semantics-for-logical-udt
• gestalt condensation-correspondence-theorem — informs 10 goals; the derived number multiplies 5 unread guesses; your holistic number for condensation-correspondence-theorem (about two minutes) sets its promise — under condensation-correspondence-theorem
• gestalt tt-finite-frames-settled — informs 1 goal; the derived number multiplies 3 unread guesses; your holistic number for tt-finite-frames-settled (about two minutes) sets its promise — under tt-finite-frames-settled
• gestalt udt-trust-existence-and-tightness — informs 1 goal; the derived number multiplies 3 unread guesses; your holistic number for udt-trust-existence-and-tightness (about two minutes) sets its promise — under udt-trust-existence-and-tightness
• gestalt ddb-finite-frames-formalized — informs 9 goals; the derived number multiplies 4 unread guesses; your holistic number for ddb-finite-frames-formalized (about two minutes) sets its promise — under ddb-finite-frames-formalized
• gestalt timely-pointwise-tower-refuted-in-setting — informs 2 goals; the derived number multiplies 4 unread guesses; your holistic number for timely-pointwise-tower-refuted-in-setting (about two minutes) sets its promise — under timely-pointwise-tower-refuted-in-setting
• gestalt mart-implies-value-refuted-unconditional — informs 1 goal; the derived number multiplies 4 unread guesses; your holistic number for mart-implies-value-refuted-unconditional (about two minutes) sets its promise — under mart-implies-value-refuted-unconditional
• gestalt inferability-and-positivity — informs 1 goal; the derived number multiplies 4 unread guesses; your holistic number for inferability-and-positivity (about two minutes) sets its promise — under inferability-and-positivity
• gestalt selection-collapse--corrigibility-risk — informs 1 goal; the derived number multiplies 4 unread guesses; your holistic number for selection-collapse--corrigibility-risk (about two minutes) sets its promise — under selection-collapse--corrigibility-risk
• gestalt anticipation-branchwise-dominance--corrigibility-risk — informs 1 goal; the derived number multiplies 4 unread guesses; your holistic number for anticipation-branchwise-dominance--corrigibility-risk (about two minutes) sets its promise — under anticipation-branchwise-dominance--corrigibility-risk
• gestalt conditioning-manages-news — informs 1 goal; the derived number multiplies 4 unread guesses; your holistic number for conditioning-manages-news (about two minutes) sets its promise — under conditioning-manages-news
• gestalt kernel-conditioning-fails-witness — informs 2 goals; the derived number multiplies 4 unread guesses; your holistic number for kernel-conditioning-fails-witness (about two minutes) sets its promise — under kernel-conditioning-fails-witness
• gestalt ps-cor614-orbit-count — informs 7 goals; the derived number multiplies 4 unread guesses; your holistic number for ps-cor614-orbit-count (about two minutes) sets its promise — under ps-cor614-orbit-count
• gestalt shutdown-button-toy — informs 7 goals; the derived number multiplies 4 unread guesses; your holistic number for shutdown-button-toy (about two minutes) sets its promise — under shutdown-button-toy
• gestalt reflective-oracles-correlated — informs 1 goal; the derived number multiplies 4 unread guesses; your holistic number for reflective-oracles-correlated (about two minutes) sets its promise — under reflective-oracles-correlated
• gestalt base-rate-working-picture — informs 1 goal; the derived number multiplies 4 unread guesses; your holistic number for base-rate-working-picture (about two minutes) sets its promise — under base-rate-working-picture
• gestalt generalized-lic — informs 1 goal; the derived number multiplies 4 unread guesses; your holistic number for generalized-lic (about two minutes) sets its promise — under generalized-lic
• gestalt local-criterion-three-cells — informs 1 goal; the derived number multiplies 4 unread guesses; your holistic number for local-criterion-three-cells (about two minutes) sets its promise — under local-criterion-three-cells
• gestalt exact-self-trust-refuted — informs 1 goal; the derived number multiplies 4 unread guesses; your holistic number for exact-self-trust-refuted (about two minutes) sets its promise — under exact-self-trust-refuted
• gestalt general-modification-object — informs 1 goal; the derived number multiplies 4 unread guesses; your holistic number for general-modification-object (about two minutes) sets its promise — under general-modification-object
• gestalt void-branch-deterrence-floor--corrigibility-risk — informs 6 goals; the derived number multiplies 4 unread guesses; your holistic number for void-branch-deterrence-floor--corrigibility-risk (about two minutes) sets its promise — under void-branch-deterrence-floor--corrigibility-risk
• gestalt li-corrigibilization-theorem — informs 1 goal; the derived number multiplies 4 unread guesses; your holistic number for li-corrigibilization-theorem (about two minutes) sets its promise — under li-corrigibilization-theorem
• gestalt fa-trust-half-weak-grades — informs 9 goals; the derived number multiplies 6 unread guesses; your holistic number for fa-trust-half-weak-grades (about two minutes) sets its promise — under fa-trust-half-weak-grades
• gestalt condensation-scores-explained — informs 1 goal; the derived number multiplies 3 unread guesses; your holistic number for condensation-scores-explained (about two minutes) sets its promise — under condensation-scores-explained
• gestalt basin-formal-object — informs 2 goals; the derived number multiplies 4 unread guesses; your holistic number for basin-formal-object (about two minutes) sets its promise — under basin-formal-object
• gestalt corrigibility-anti-natural — informs 1 goal; the derived number multiplies 3 unread guesses; your holistic number for corrigibility-anti-natural (about two minutes) sets its promise — under corrigibility-anti-natural
• gestalt mm-results-in-li — informs 1 goal; the derived number multiplies 5 unread guesses; your holistic number for mm-results-in-li (about two minutes) sets its promise — under mm-results-in-li
• gestalt cleanroom-deference-followups — informs 1 goal; the derived number multiplies 3 unread guesses; your holistic number for cleanroom-deference-followups (about two minutes) sets its promise — under cleanroom-deference-followups
• gestalt hedged-value-unconditional — informs 3 goals; the derived number multiplies 5 unread guesses; your holistic number for hedged-value-unconditional (about two minutes) sets its promise — under hedged-value-unconditional
• gestalt agency-correspondence-refuted — informs 1 goal; the derived number multiplies 3 unread guesses; your holistic number for agency-correspondence-refuted (about two minutes) sets its promise — under agency-correspondence-refuted
• gestalt ro-generalized-kakutani-refuted — informs 1 goal; the derived number multiplies 3 unread guesses; your holistic number for ro-generalized-kakutani-refuted (about two minutes) sets its promise — under ro-generalized-kakutani-refuted
• … 240 more unranked units in derived.json (--all prints them)