Skip to content

probe(hebb): saturation-damped local plasticity versus episodic MRU baseline - #1722

Merged
AdaWorldAPI merged 1 commit into
mainfrom
lab/ce64-hebb-saturation-probe-20261022
Oct 22, 2026
Merged

AdaWorldAPI merged 1 commit into
mainfrom
lab/ce64-hebb-saturation-probe-20261022

Conversation

@AdaWorldAPI

Copy link
Copy Markdown
Owner

Purpose

Isolate Google's idea of a damped Hebbian update Δz = eta * s * p * (1 - tanh(abs(z))) as an experimental candidate, not a rewrite of NARS F/C or an instruction to modify immutable Palette256.

Current architecture

crates/lance-graph-contract/src/episodic_edges.rs already encodes the hot episodic promote (MRU order) and explicitly leaves numerical plasticity to the co-addressed CausalEdge64. A separate owner must decide storage/learning contracts. No new register bits.

New independent falsifiers

  • Inertness: learning rate=0 and either activation=0 leave z unchanged.
  • Can-fire: coactivation at z=0 yields positive increase; opposing activation yields negative change.
  • Can-stay-silent: equal stimuli at high |z| give smaller updates than at z=0, yet no source data or immutable codebook is touched.
  • A repeated 64-step high-activity sequence grows more slowly than undamped linear Hebbian accumulation.
  • Existing episodic MRU ordering remains independent of the suggested weight update.

Important boundaries

The scalar formula is a candidate learning policy, NOT certified Bayesian updating, NOT a proven anti-hallucination mechanism, and NOT a production bitfield writer. The numerical update is evaluated only in this test. Domain and effect state are explicit. Await D-CGW/D-HL owner admission and compare against independent unsaturated baseline before integration.

Verification

cargo test --manifest-path crates/lance-graph-contract/Cargo.toml --test episodic_hebb_saturation_probe.
Lab test only, no production code changed.

@coderabbitai

coderabbitai Bot commented Oct 22, 2026 •

Copy link
Copy Markdown

Important

  • 🔍 Trigger review

This repository does not receive automatic reviews because it has fewer than 10 stars.

⚙️ Run configuration
  • Source mode: Auto
  • Repository: Unavailable
  • Branch: Unavailable
  • File: No configuration file from this repository applied
  • Commit: Unavailable
  • Layers: Organization UI
  • Review profile: CHILL
  • Plan: Advanced
  • Run ID: b5751d15-9c66-4afd-937c-4ee96c893c58
  • Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

Warning

Billing warning: we have not been able to collect payment for this subscription for more than 72 hours. Please update the payment method or pay any pending invoices in Billing to avoid service interruption.

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

Comment @coderabbitai help to get the list of available commands.

AdaWorldAPI pushed a commit that referenced this pull request Oct 22, 2026
Moves crates/lance-graph-contract/tests/episodic_hebb_saturation_probe.rs here from #1722. Unlike the local helpers already in this
series, it exercises the real implementation, so the series keeps its
experiment contract and gains the concrete counter-probe. #1722 is
closed as superseded once this lands.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01G22yT6htkcdyXsihxxXdrg

Copy link
Copy Markdown
Owner Author

This test now lives in #1713 (commit b3d26a4), where it serves as the production-surface counter-probe for the canonical #1709–#1717 series. It passes locally (release). I'll close this PR as superseded once #1713 has landed.


Generated by Claude Code

@AdaWorldAPI
AdaWorldAPI marked this pull request as ready for review October 22, 2026 21:55
@chatgpt-codex-connector

Copy link
Copy Markdown

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review Status Commit Review trigger
📝 Code Review 🔄 Running since 2026-10-22T21:55:38.393949Z c687758 Draft marked ready
ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

@AdaWorldAPI
AdaWorldAPI merged commit 2cc7aa0 into main Oct 22, 2026
20 checks passed
AdaWorldAPI added a commit that referenced this pull request Oct 22, 2026
…nce-20261022

board: coresearch map for the Palette256 evidence/plasticity probes (#1706–#1722)
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant