Bump actions/deploy-pages from 4 to 5#6
Open
dependabot[bot] wants to merge 1 commit into
Open
Conversation
Bumps [actions/deploy-pages](https://github.com/actions/deploy-pages) from 4 to 5. - [Release notes](https://github.com/actions/deploy-pages/releases) - [Commits](actions/deploy-pages@v4...v5) --- updated-dependencies: - dependency-name: actions/deploy-pages dependency-version: '5' dependency-type: direct:production update-type: version-update:semver-major ... Signed-off-by: dependabot[bot] <support@github.com>
dependabot
Bot
added
dependencies
dependabot
Bot
added
dependencies
timm156
added a commit
that referenced
this pull request
Jul 7, 2026
…-001 fixed, T-001 provenance flagged O-001 (BLOCKER #4 RESOLVED — best case, number vindicated): The 473,622:1 octahedral headline was null in O-001's own results.json because the n=4 handler post-dates its 2026-03-18 run — not because the number is wrong. Re-derived from the saved bridge tensors with the exact train-time metric (train_exp2_scale.coplanar_crossplanar_ratio + train_cybernetic.py:581-589 aggregation): - FO-001 -> 262,920.298 (bit-exact to its logged value — validates the method) - O-001 -> 473,621.655 (== 473,622 rounded — vindicated) Persisted as a re-runnable anchor: scripts/rederive_octahedral_cocross.py + results/octahedral-hermes-anchor/{FO-001,O-001}/bridges/ (tensors force-added past the .npy gitignore) + co_cross_rederivation.json. Both n=4 endpoints of the inverse-n ordering (474K>70K>42K) now rest on reproducible local evidence. BM-001 (BLOCKER #6 RESOLVED): results/BM-001/RESULTS.md aggregate Mean cells corrected 0.6970/0.6982 -> true 4-primary-metric means 0.72205/0.723225 (Base 0.7417 reproduces). Delta +0.0012 unchanged (the ~0.025 offset cancels). Per-benchmark table was exact. T-001 (BLOCKER #5 — reclassified OPEN, NEEDS PI; a premature edit reverted): The Director asked me to reconcile "which metric / which run-pairing". Chasing it surfaced a bigger problem than the stale "34/6 checkpoints / 3.5%": the paper's r=1.0000 reproducibility compares T-001r1 (2,700 steps, 5,395:1 @2700) vs T-001r2, and the T-001r1 (2,700) artifact is NOT on local disk. The only local candidate, T-001-full (7,100 steps), is a DIFFERENT run (5,018:1 @2700, ~7% off); T-001-full-r2 matches r2 (41,564 @10k exact). My earlier-in-session "correction" (co/cross r=0.99836 / Fiedler 0.99995, from the T-001-full vs T-001-full-r2 pair — which the Director and I agree on) was the WRONG pairing for the paper's claim; I caught it (5,018 != 5,395) and reverted it before it shipped. The paper's T-001 numbers are left untouched and flagged: PI must locate the canonical r1(2,700)+r2 artifacts (Hermes / renamed / re-run) before they can be verified. A second unbacked-headline, in kind with O-001. Header blockers updated (#4/#6 RESOLVED, #5 OPEN-NEEDS-PI); no claim shipped. PAPER4_VERIFICATION_PASS.md rows + narrative updated to match. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01JQ1bCmAJ5PdNb8q8NboQy3
timm156
added a commit
that referenced
this pull request
Jul 7, 2026
…r reframe Both adversarially verified in two rounds (a 3-agent verify workflow + a focused re-check); every number confirmed against on-disk data. BLOCKER #2 (P0 false attribution) — RESOLVED. train_comparison.py is pure LM loss (loss=outputs.loss; zero spectral/contrastive/Steersman code). section8: - Overhead reattributed to training the learnable bridge (the ONLY config difference from the standard-LoRA baseline, which carries the same n=6 bridge frozen at identity). Dropped the wrong "control law evaluation" AND the round-1 wrong substitute "per-step Fiedler diagnostics" (Fiedler is logged only every 500 steps, in both arms). Fig caption fixed likewise. - The adjacent "evolves under spectral pressure / spectral attractor" claim (also false — no spectral loss) recast as a pure-LM near-identity baseline. - Minors: loss "identical to 4 sig figs (0.01%)" -> "3 sig figs (~0.06%)"; "defaults to the spectral attractor" -> "near-initialization resting state". BLOCKER #3 (spectral-attractor reframe + H-ch4) — RESOLVED. - H-ch4 Fiedler 0.0836 (a 1,100-step PARTIAL; local run stops at 1100) -> 0.0918 (10,000-step final). Anchored with H-ch8 0.0944 from Hermes at results/attractor-hermes-anchor/ (neither was cleanly on local disk). Band 0.0836-0.1019 -> 0.0918-0.1019; CoV 10.4% -> 3.9% (population). Fixed every site incl. the tab:hierarchy (section6) and taxonomy (section7) bands rounded to "0.084", while preserving E-001's own distinct Fiedler 0.084. - Reframe: BM-000 places Fiedler ~0.09 inside the identity+noise (eps=0.05) near-init null (~17th pct). Recast across abstract, intro, section5, section7, section8, discussion as a near-initialization resting state, NOT a structured "attractor/equilibrium". BM-000 now described inline in section5 (self-reference removed). The bifurcation (contrastive Fiedler far below every null, gauss6 max 0.0136) strengthened as the null-validated result. Also: flagged the stale parallel build (rhombic-paper4.tex -> paper4-sections/, last touched March, no reframe) with a DO-NOT-COMPILE header; paper4-main.tex is canonical. Header records two out-of-scope follow-ups (dual regime-numbering, fuller BM-000 methods) for the-adversary rewrite round. Blocker header now: #1/#2/#4/#6 RESOLVED, #3 RESOLVED, #5 OPEN-NEEDS-PI (T-001). Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01JQ1bCmAJ5PdNb8q8NboQy3
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.
Bumps actions/deploy-pages from 4 to 5.
Release notes
Sourced from actions/deploy-pages's releases.
... (truncated)
Commits
cd2ce8fMerge pull request #404 from salmanmkc/node24bbe2a95Update Node.js version to 24.x854d7aaMerge pull request #374 from actions/Jcambass-patch-1306bb81Add workflow file for publishing releases to immutable action packageb742728Merge pull request #360 from actions/dependabot/npm_and_yarn/npm_and_yarn-513...7273294Bump braces in the npm_and_yarn group across 1 directory963791fMerge pull request #361 from actions/dependabot-friendly51bb29dMake the rebuild dist workflow safer for Dependabot89f3d10Merge pull request #358 from actions/dependabot/npm_and_yarn/non-breaking-cha...bce7355Merge branch 'main' into dependabot/npm_and_yarn/non-breaking-changes-99c12deb21Dependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting
@dependabot rebase.Dependabot commands and options
You can trigger Dependabot actions by commenting on this PR:
@dependabot rebasewill rebase this PR@dependabot recreatewill recreate this PR, overwriting any edits that have been made to it@dependabot show <dependency name> ignore conditionswill show all of the ignore conditions of the specified dependency@dependabot ignore this major versionwill close this PR and stop Dependabot creating any more for this major version (unless you reopen the PR or upgrade to it yourself)@dependabot ignore this minor versionwill close this PR and stop Dependabot creating any more for this minor version (unless you reopen the PR or upgrade to it yourself)@dependabot ignore this dependencywill close this PR and stop Dependabot creating any more for this dependency (unless you reopen the PR or upgrade to it yourself)