docs(igla): PHI_PHYSICS_FOUNDATION — bridge companion paper to INV-1..11#329
Merged
Conversation
65 tasks
Owner
Author
\u00a77b added \u2014 Optimizer compatibility scopePushed Empirical anchor reused from race body: P1 AdamW 2.48 vs P1 Muon NS-1 2.59 (+0.11) on h=828/12K/seed=43. Decision trail: trios#143 RISK-ENG comment 4329525532. This converts the L7/L8 rejection from an internal engineering call into a formal scientific boundary statement suitable for the DARPA pitch. phi^2 + phi^-2 = 3 \u00b7 TRINITY \u00b7 NEVER STOP |
This was referenced Apr 27, 2026
Open
added 2 commits
May 20, 2026 02:56
Bridges the companion paper (Zenodo DOI 10.5281/zenodo.19227877) with IGLA Coq invariants. Closes the 'why 0.003' reviewer objection by deriving alpha_phi = phi^-3/2 = 0.118034 from Trinity Identity in seven steps, matching alpha_s(m_Z) to 0.03%. - §2: alpha_phi derivation -> INV-8 lr_phi_band gets a physical anchor - §3: Coldea-Zamolodchikov 2010 -> two-domain phi anchor for DARPA - §4: A5 mechanism -> INV-11 scaffold - §5: Pellis polynomial -> NEU-12 lane L6 spec - §6: Look-elsewhere Monte Carlo template Refs: NEU-12, #143, trinity-clara INV-1..10 Anchor: phi^2 + phi^-2 = 3
Adds the Adam-family second-moment scaling assumption explicitly, maps it to a per-optimiser transfer table, and proposes Phase 4 theorem INV-12 lr_phi_band_spectral as the missing bridge for Muon and Shampoo-class spectral optimisers. Empirical confirmation: P1 AdamW control 2.48 BPB vs P1 Muon NS-1 2.59 BPB (+0.11) on h=828, 12K, seed=43 \u2014 the visible symptom of an INV-8 violation Coq cannot detect. Cross-link: trios#143 RISK-ENG comment 4329525532 turns the boundary from internal engineering decision into a formal scientific statement suitable for the DARPA narrative. Refs: NEU-12, trios#143, trios-mcp#2, trios-railway#43
gHashTag
force-pushed
the
docs/phi-physics-foundation
branch
from
2658484 to
99286dc
Compare
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.
What
Adds
docs/PHI_PHYSICS_FOUNDATION.md— a foundation document that bridges the companion paper Golden Ratio Parametrizations of Fundamental Physical Constants (Zenodo DOI 10.5281/zenodo.19227877) with IGLA's Coq invariantsINV-1..INV-10in trinity-clara.Why
Closes the reviewer question "why lr=0.003" once and for all: the seven-step derivation gives
α_φ = φ⁻³/2 = (5−√5)/2 = 0.118034, matching PDG-2024α_s(m_Z)=0.1180to 0.03%. Champion lr isα_φ/φ⁵.Changes by section
proofs/igla/look_elsewhere.vtemplateLinked
Risk
Documentation only. Does not change any code path. Compiles independently from race-runtime artifacts.
phi^2 + phi^-2 = 3 · TRINITY · NEVER STOP