You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
feat(tri): math compare --weinberg (φ⁻³ golden test) #295
Add tri math compare --weinberg (or equivalent t27c flag) to evaluate row 22: φ⁻³ vs PDG sin²θ_W with an explicit tolerance and JSONL fields for reproducibility.
Motivation
FORMULA_TABLE.md row 22 and TRINITY_VS_SM_FORMULAS.md §4 document the conjectural ansatz sin²θ_W ≈ φ⁻³ (~0.236068 vs PDG central ~0.23122, ~2.1%).
Falsifiability path: P2@MESA / DUNE ND (see FORMULA_TABLE.md Mechanism note).
Today there is no dedicated CLI path; golden test would lock Δ under a declared bound (e.g. < 0.005 on |φ⁻³ − sin²θ_W| only after PDG reference + scheme are frozen in spec).
Acceptance criteria
CLI flag documented next to --hybrid-v2 / --pellis-extended in bootstrap/src/math_compare.rs (or factored module).
Reference sin²θ_W from a single documented source (PDG / preset constant in pellis-formulas.t27 or sibling spec) — no silent hard-coded magic without spec pointer.
Print φ⁻³, reference sin²θ_W, absolute and relative Δ; append JSONL keys e.g. weinberg_enabled, phi_inv_cubed, sin2_theta_w_ref, delta_abs, delta_rel.
#[cfg(test)]: golden test |φ⁻³ − sin²θ_W_ref| < 0.005 (adjust if scheme requires).
English docs: one paragraph in FORMULA_TABLE.md row 22 status or hybrid-conjecture.md pointing to the flag.
Non-goals
Claiming H1 or electroweak “proof”; this is a diagnostic / conjecture hook only.
Replacing PDG MS-bar running with a full renormalization story (document scheme in spec).
Activity
docs(pellis): RU integration table, PR #325 row, dedupe issues note
docs(pellis): RU integration table + dedupe note (without stray specs)
docs(pellis): WORK_REPORT §1 — full EN timeline, PR #328, no RU block
docs(pellis): RU integration table, PR #325 row, dedupe issues note
docs(pellis): RU integration table + dedupe note (without stray specs)
docs(pellis): WORK_REPORT §1 — full EN timeline, PR #328, no RU block
gHashTag commentedon Apr 8, 2026
📓 NotebookLM Notebook created
Notebook ID:
e44f00c2-897e-4e4f-bd7d-b40295c5c076URL: https://notebooklm.google.com/notebook/e44f00c2-897e-4e4f-bd7d-b40295c5c076
gHashTag commentedon Apr 29, 2026
Duplicate of #300 (same feature: tri math compare --weinberg).
docs(pellis): RU integration table, PR #325 row, dedupe issues note
docs(pellis): RU integration table + dedupe note (without stray specs)
docs(pellis): WORK_REPORT §1 — full EN timeline, PR #328, no RU block
docs(pellis): RU integration table, PR #325 row, dedupe issues note
docs(pellis): RU integration table + dedupe note (without stray specs)
docs(pellis): WORK_REPORT §1 — full EN timeline, PR #328, no RU block