docs(verdicts): RIDE displacement arm separated-negative - residual closed on evidence - #234
Merged
Merged
Conversation
added 4 commits
October 2, 2026 09:43
…losed on evidence
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Verdict for the RIDE displacement arm (TODO.sota-2026/05, executed after "Proceed all").
Result: SEPARATED-NEGATIVE - run-016-ride scores 5.8627 DER-CE (full 1,200 paragraphs), +1.3926pp [1.156, 1.646] p=0.0 worse than the 2.1 rung (4.5701) under the paired between-students bootstrap. Teacher reproduced at 2.2921. Training was healthy (CE 0.57 -> 0.29, 13,026 steps) and the loss budget was pre-registered (beta auto-calibrated to 10% of CE at start).
This is the strongest test of the student-side closure rule to date: the arm is the only one in the ledger whose premise was measured BEFORE training (the direction-transferability probe passed decisively - max-layer cos 0.9513), and it still failed at the outcome. Reading: a domain-general residual direction is necessary but not sufficient; regressing a 300M byte student's encoder toward ridge-projected 580M targets displaces representations the decoder relies on, competing with the CE objective rather than sharpening it.
Changes: TODO 05 closed with the measured result; RESULTS.md ledger row 9 + "closed on evidence, not exhaustion" entry; paper.adoc residual paragraph gains the representation-space arm sentence. The frontier mover remains teacher-side only - run-009-yallamorph is in flight (55%+).