Proprio-as-conductance is refused: the shunt that smooths the wave kills the reversal - #210
Merged
Merged
Conversation
…lls the reversal The channel translation of the proprioceptive drive won every guardrail it was originally measured on -- and every number in that table was a forward-gait number. The behavioural battery (paired on identical seeds, three assay families, issue #194) measured what the guardrails never did: spontaneous reversals in clean space fall from 2.03/min to exactly zero, the off-food ethogram produced 1 reversal event against the baseline's 9 with reorientation collapsing 53 to 13 degrees, chemotaxis approach ends 15.4 mm worse, and the pirouette food-conditioning ratio inverts. The mechanism reading is recorded with the tables: a conductance does not just drive, it SHUNTS. Both halves raise the cord's total membrane conductance, so when the backward command tries to take the cord over it is pushing on a leakier membrane and loses its lever. The wave keeps its shape and loses its ability to change direction -- and with H2 already null (the cords never occupied the rail), the translation had no problem to fix and a real behaviour to cost. Current mode stays the animal. The refusal lives in tools/clamp_occupancy.py beside the sweep that motivated the candidate, and sensory.proprio_conductance is now pinned in RUNTIME_UNSUPPORTED -- it was a Python-only default with no registry entry, which was a parity gap regardless of the verdict. The AS-field registry note is updated to its own refusal (#193) in passing. Closes #194. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CBroMXcYdfkyb5buJPcKvL
YesterdaysLemon
marked this pull request as ready for review
August 27, 2026 01:31
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
2 participants
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.
Closes #194.
The candidate
The channel translation of the proprioceptive drive (
SensoryParams.proprio_conductance, 5 nS) won every guardrail it was originally measured on — dv_corr −0.73 → −0.82, TWI +0.87 → +0.91, speed 0.304 → 0.361 mm/s. Every number in that table was a forward-gait number.The battery (paired on identical seeds, three assay families)
The mechanism reading
A conductance doesn't just drive, it shunts. Both halves raise the cord's total membrane conductance, so when the backward command tries to take the cord over, it pushes on a leakier membrane and loses its lever. The wave keeps its shape and loses its ability to change direction. With H2 already null (the cords never occupied the rail — the census in #206), the translation had no problem to fix and a real behaviour to cost.
What changed
tools/clamp_occupancy.pycarries THE BATTERY beside the sweep that motivated the candidate.sensory.proprio_conductanceis now pinned inRUNTIME_UNSUPPORTED— it was a Python-only default with no registry entry, a parity gap regardless of verdict (proprio_E_revneeds no entry: never read at zero, theload_halfshape).No model change — current mode was already the shipped animal; this PR is the record of why it stays.
With this, both Track A adoption candidates from the wake-up menu (#193, #194) are measured and refused, each with a mechanism lesson. The dossiers were worth writing: the refusals are as informative as the adoptions.
🤖 Generated with Claude Code
https://claude.ai/code/session_01CBroMXcYdfkyb5buJPcKvL
Generated by Claude Code