Buffer has no second attractor: the flip is an episode, and PVD feels the coil - #213
Merged
Merged
Conversation
… the coil The scorecard's buffer row said one seed in five settles into a 0.33 Hz whole-body standing flip. Watching instead of sampling says otherwise (tools/buffer_basin.py, issue #195): twelve seeds over two minutes, eleven swim cleanly; the one deep excursion collapses the travelling index to 0.08 and then ENDS BY ITSELF, back to a clean wave by 135 s and holding for the next three minutes. The pre-PVD model shows the same shape with different timing -- episodes at 45-60, 105-120, 165-180 s, every one self-ending. The 2026-08-25 scorecard reading was one such episode straddled by its own 46-86 s analyse window: a snapshot of a spell, reported as a stationary state. The hunt also caught something better than the answer it went in for. The buffer trajectories of the current and pre-PVD models fork at t~55 s -- bit-identical for 11,000 steps, then different -- and the bisect lands on the PVD merge. The cause is honest: the contact total PVD reads includes Body.self_contact_force, and a swimming animal's deep coil presses its own far side at whole-body totals of 3.6-8.3, past pvd_threshold = 3.0. A hard coil is now a nociceptive event. Biologically defensible -- PVD tiles the body wall -- and now on the record in the PVD dossier rather than implicit in a trajectory diff: agar and viscous never self-contact at these totals, which is why every determinism pin held while buffer moved. Bookkeeping: the scorecard's bimodality note (print and doc) now describes the episode instead of asserting a basin, the PVD calibration dossier carries the self-contact regime, and NEXT.md item 1's protocol guard points at the map. Closes #195. 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 05:56
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
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 #195.
The map
The scorecard's buffer row said "one seed in five settles into a ~0.33 Hz whole-body standing flip." Watching instead of sampling (
tools/buffer_basin.py, 15 s windows over long swims):Buffer has no second attractor at shipped defaults. It has a recurrent ~15 s coil-up excursion the animal always exits.
The better find
Current and pre-PVD buffer trajectories are bit-identical for 11,000 steps and then fork at t≈55 s; the bisect lands on the PVD merge (#205). Cause: the contact total PVD reads includes
Body.self_contact_force, and a swimming animal's deep coil presses its own far side at whole-body totals of 3.6–8.3 — pastpvd_threshold = 3.0. A hard coil is now a nociceptive event. Biologically defensible (PVD tiles the body wall; harsh body contact is harsh body contact), and now on the record in the PVD dossier rather than implicit in a trajectory diff. Agar and viscous crawls never self-contact at these totals — which is exactly why every determinism pin stayed green while buffer moved.What changed
tools/buffer_basin.py— the instrument, with THE RECORD.params.py) carries the self-contact regime, flagged as a mechanism decision with evidence if self-touch should ever be excluded.No model change — instrument, record, and corrected prose only.
🤖 Generated with Claude Code
https://claude.ai/code/session_01CBroMXcYdfkyb5buJPcKvL
Generated by Claude Code