training-map-model
leaf/question · drafted
Statement
A finite model of the training map, from the overseers' data and loss to the next model's state with a measure of what the next model retains, exists and reproduces the overwriting-dodge verdict inside it.
Source
projects/deference/faf-cleanroom-2026-10-02/sources/corrigibility/workflow-2026-09-14b/plan/synthesis.md, summary and agenda N2 ("Both name a model of the training map T with a retention measure as the most valuable missing work"); projects/deference/faf-cleanroom-2026-10-02/sources/corrigibility/workflow-2026-09-14b/develop/legitimacy-general-final.md; the clean-room result y1_dodge in projects/deference/faf-cleanroom-2026-10-02/packages/corr-legit-modif, which the model is to re-run.
Notes
The object both run-3 planners rank first; the overwriting-dodge verdict and the values-ahead condition both need it.
Author and human events
Author @smithy-verity; version 1; size L.
No human event has touched this node.
Edges
Above
- part of joint-process-model
Ranking (the lens, not the structure)
{
"status": "open",
"diagnosis": null,
"flags": [
"AI number"
],
"V": {
"V_prior": null,
"display": "2",
"headline": 2,
"ref": 2,
"valid": 2,
"weaker": "both"
},
"belief": {
"J": 0,
"mid": "23/40",
"range": [
"3/40",
"1/1"
],
"width": "1/1"
},
"draft": "ai"
}Say something about it
Each verb opens a prefilled issue; a round files what lands as the ledger event it is.