Single Source of Truth — cross-paper dashboard

SSOT & Task Queue

Canonical arXiv-readiness and close-the-gap work for all four BigBounce papers. Every number on this page is sourced from the per-paper status.md files in project-context/SSOT/. When surfaces elsewhere on the site disagree with this page, this page wins.

Last authoritative update: 2026-04-22 (fire #189) · 🟢 Phase 2 Path-C rebuild of Paper 3 CLOSED — 12/12 criteria green, weighted sum 100.000 % · drive-to-100 loop self-terminated · view raw SSOT on GitHub

🟢 Paper 3 — Path-C Rebuild CLOSED (fire #189, 12/12, 100 %)

All four papers now at 100 % on content + compile axes. Paper 3 Path-C rebuild (initiated 2026-04-19 after Houston novelty-integrity pushback) CLOSED 2026-04-22 fire #189 with all 12 exit criteria green. Tracked in SSOT/drive-to-100.md. Final state: 12 CLOSED — #1 SDSS native re-score (1,925,279 scored / ~6500× anomaly-rate reduction, fire #164); #2 LAMOST native re-score (21.4× reduction); #3 CMB native retrain (500/500 = 100 % injection-recovery); #4 DESI 5-fold k-fold stability (fire #189: live SPARCL 47,000 spectra, Jaccard gate PASS at mean pairwise &Jbar; = 0.862 vs required ≥ 0.70, consensus ≥3/5 = 85.0 % of union, DESI headline 195,829 unchanged); #5 NEOWISE ecliptic mask; #6 injection-recovery 3 plant morphologies × 4 surveys; #7 7-way dedup 378,280 unique / 637 multi-survey (fire #164; ACT excluded R42-fix); #8 Paper 3 integration (fire #167); #9 PDF recompile 28 MB / 33 pp / 0 undef (fire #168, re-mirrored fire #189 with k-fold paragraph); #10 HF rebuild 5-file 15.2 MB (fire #166); #11 P1-PDF-V3 14 bundled edits → 975 KB / 27 pp / 0 undef; #12 site-sync. Weighted sum 12.000 / 12 = 100.000 %. Cron 9f44c29e self-terminated. Pod retained for Houston inspection; cumulative Path-C spend ~$71 of $400 cap.

Recently completed

fire #51 — Paper 4 post-TTA 0.43σ generator added (P4-POST-TTA-043SIGMA-TRACE): added pipelines/p2_chirality/run_dipole_catalog_c.py (241-line pod-runnable generator) that reproduces the paper's headline 0.43σ null-dipole by applying the same HEALPix NSIDE=64 / min-10/pixel / hp.fit_dipole / 1,000 MC-null pipeline from run_dipole_8M.py (Catalogs A/B) to Catalog C's catalog_production.parquet (post-TTA equivariant class_eq labels). JSON output is audit-grade — encodes paper claim + 8 tex line-numbers + methodology block + full provenance chain. Closes the dipole-skeptic "add generator or retract" gap without paper edits. Last truly agent-doable row in the queue; all remaining rows pod- or Houston-gated.
fire #50 — CLAUDE.md stat cards triple-verified (P-CLAUDEMD-NANOGRAV-QUINTOM-TRIPLE-VERIFY): audited ~25 "Key scientific results" bullets against the 4 committed paper .tex sources. 2 discrepancies caught: (1) line 45 "33.5M sources" → 37.3M (Paper 3 Table 1 canonical total 37,292,042); (2) line 59 PBH bullet stale γ=3.33±0.40 / 0.83σ → harmonized to γ=3.20±0.42 / 0.48σ matching Paper 3 §6. All remaining paper-anchored claims verified verbatim (424,781/309,789 MCMC · H₀=67.68 · β=0.27°/0.342° · f_NL=-35/8 · per-survey anomaly counts · 8.47M galaxies). All agent-doable P0/P1 rows now closed — remaining 6 rows are Houston-owned or pod-owned.
fire #49 — Paper 4 MC realization count harmonized (P4-MC-REALIZATIONS-HARMONIZE): all four in-text 10,000 MC-null mentions changed to 1,000 matching dipolar_analysis.log production run. New footnote cites the log + notes the 3% tail-$p$ precision is immaterial at non-$3\sigma$ Catalog~C significances. Methodological-metadata fix, central statistics unchanged. PDF 25,743,193→25,764,622 B, tarball 20.15 MB, both smoke-tested
fire #48 — Paper 1 sample-count footnote added (P1-SAMPLE-COUNT-FOOTNOTE): §IV.A footnote at arxiv/main.tex L328 reconciles 424,781 / 309,789 / 176,240 / 123,129 / 119,617 sample counts without retraction — all 4 numbers trace to the same chains via stratification (full sum → 2 frozen combos → full-tension alone → post-burnin → getdist-thinned). PDF 945,753→965,999 B, 0 undef cite/ref, 27 pages
fire #47 — Paper 4 raw-axis hidden positive disclosed (P4-RAW-AXIS-18P9DEG-DISCLOSE): new paragraph at §VIII.A surfaces that the pre-TTA raw dipole axis is 18.9° from Shamir's claimed axis (aligned) and fully decorrelates post-TTA — strengthens the null-dipole thesis by showing the equivariance correction neutralizes both amplitude AND direction of the systematic. PDF 25,743,193 B (+879 B), tarball 20.15 MB, both smoke-tested byte-exact
fire #46 — Paper 3 Table 1 threshold disclosure (P3-TABLE1-THRESHOLD-AUDIT Major): caption rewrite enumerates the 2 threshold families ($S>5.0$ for 3 spec surveys; top-1% for Planck/ACT/Gaia/NEOWISE; top-298 cap for eROSITA). PDF 28,147,041 B, tarball rebuilt 27.67 MB, both smoke-tested
fire #45 — P3-TIC-374313355-RETRACT closed as MOOT: on-disk SIMBAD evidence at pipelines/p3_anomaly_engine/p3a_tess_374313355_lomb_scargle/typing/simbad_hit.json confirms M9V (2MASS J10403573+0505297, otype LM*, 0.20″, bibcode 2009ApJS..182..543A); no retraction needed
fire #44 — Paper 2 cites Paper 3 (P2-CITE-PAPER-3 Tier-A): added Golden:2026anomaly bibitem + 3 \cite{} insertions in §4 SPHEREx, §5 MegaMapper, §7.2 PNG-Bias; recompile clean, PDF 633,183 B, tarball rebuilt 320 KB
fire #43 — Paper 3 arXiv tarball built (27.6 MB, 23 files: tex + 22 figures); 28,146,437 B PDF reproduced exactly, 0 undef refs, inline thebibliography
fire #42 — Paper 2 arXiv tarball rebuilt with embedded .bbl (319,960 B, 9 files); 632,145 B PDF reproduced exactly, 0 undef refs
fire #41 — all 4 papers recompiled via Docker TeX Live (P1 945 KB · P2 632 KB · P3 28 MB · P4 25.7 MB, 0 undef refs each) → public/papers/ refreshed (310c1d9)

Currently working

All 4 arXiv tarballs now built + smoke-tested — P1 (fire #9), P4 (fire #39, 19 MB), P2 (fire #42, 320 KB), P3 (fire #43, 27.6 MB). Houston-owned step next: arXiv upload.
Peer-review P2 followups — P1-SAMPLE-COUNT-FOOTNOTE, P2-CITE-PAPER-3, P3-TABLE1-THRESHOLD-AUDIT queued for batch fires
CLAUDE.md triple-verify — P-CLAUDEMD-NANOGRAV-QUINTOM-TRIPLE-VERIFY: re-verify stat cards against paper .tex

Next up

P-CLAUDEMD-NANOGRAV-QUINTOM-TRIPLE-VERIFY — re-verify CLAUDE.md stat cards against paper .tex after fire #25 caught 5 stale
P1-SAMPLE-COUNT-FOOTNOTE — footnote reconciling 424,781 vs 176,240 vs 119,617 vs 123,129 sample counts per Paper 1 §VII.H
P4-MC-REALIZATIONS-HARMONIZE — normalize MC realization counts across Paper 4 (10K / 1K / partial inconsistency)
P-SITE-FULL-SYNC — sweep index.html, paper.html, figures.html, data-explorer.html, glossary.html against SSOT canonical
P-HOUSTON-LSST-LINE-REVIEW — one-line review for LSST 10-yr projection in Paper 4 §6 (non-blocking for arXiv)
arXiv submission (Houston-owned) — recommended order Paper 4 → 3 → 1 → 2

Autonomous drive-to-100 loop runs every 20 min. Full fire log: project-context/SSOT/drive-to-100.md

Paper 1 — Spin-Torsion

100%

v2.3.5 · 31 pp · 63+ refs · 1.23 MB PDF (R42 Wave 11 LANDED 2026-05-01 07:30 PDT, 0 undef; abstract reframe drops "evidence for ECH"; Bayes-factor scoped to ΛCDM+ΔN_eff proxy only)

Paper 2 — f_NL Forecast

100%

v1.7.6 · 13 pp · 758 KB PDF (R42 Wave 11 LANDED 2026-05-01 07:30 PDT, 0 undef; Eq. 3 carries the 1/k² factor that surrounding prose always claimed)

Paper 3 — Anomaly Catalog

100%

v3.1.6 · 35 pp · 28.27 MB PDF (R42 Wave 11 LANDED 2026-05-01 07:30 PDT, 0 undef; retitled to 378,280 anomalies / 37.3M sources, ACT-DR6 quarantine explicit, §VI prose fix separates IF cross-check from BigAE primary catalog)

Paper 4 — Chirality Catalog

100%

v1.0.6 · 16 pp · 25.78 MB PDF (R42 Wave 11 LANDED 2026-05-01 07:30 PDT, 0 undef; N_spiral=3,321,795 shot-noise correction in footnote 5 — cross-confirmed P4-CM-B1 + P4-OA-M7; shot-noise floor up 2.55× → C_ℓ error budget up 2.65×; 3.86× equivariance suppression preserved)

Program health at a glance

#	Paper	One-line status	Ready	Canonical source
1	Spin-Torsion Cosmology	v2.3.5, 31 pp, 10+ revision rounds. R42 Wave 11 LANDED 2026-05-01 07:30 PDT — abstract reframe drops "evidence for ECH" framing; Bayes-factor scoped to "ΛCDM+ΔN_eff proxy only" (closes Wave 10 cross-model BLOCKERs P1-CM-B1 + P1-CM-B3 + P1-CM-M2 + P1-OA-B2). Recompiled on Pod 3 via texlive Ubuntu (pdflatex + bibtex + pdflatex×2): 1.23 MB PDF / 0 undef refs / theory_map figure embedded. Carries forward Wave 2/3 B1 retitle + B2 theory_map + B6 chain rerun (Rhat ~1.000, ESS ~313k).	100 %	`arxiv/main.tex`
2	f_NL Forecast (SPHEREx/MegaMapper)	v1.7.6 revtex4-2, 13 pp, 6 figures. R42 Wave 11 LANDED 2026-05-01 07:30 PDT — Eq. 3 of §III.A now carries the explicit 1/k² factor that the surrounding text always claimed (Δb(k,z) = 2·f_NL·(b₁−1)·δ_c/M(k,z) with M(k,z)=2·k²·T(k)·D(z)/(3Ω_mH₀²)) — closes claim-vs-derivation gap GPT-5 P2-OA-B1 surfaced. Recompiled on Pod 3 via texlive Ubuntu (pdflatex + bibtex + pdflatex×2): 758 KB PDF, 0 undef refs. Bayes 8–17, σ(f_NL) = 0.36 / 0.93, 1.58× Gold+Silver bias validation preserved.	100 %	`research/focused_paper_source_integration/02_full_draft.tex`
3	Multi-Survey Anomaly Catalog	v3.1.6, 35 pp. R42 Wave 11 LANDED 2026-05-01 07:30 PDT — retitled to "Spectrally Unusual Sources at Scale: A Multi-Survey Catalog of 378,280 Anomalies and Native-Trained Novelty Rates from 37.3 Million Sources" (was 319,443 anomalies / 37.3M Spectra); ACT-DR6 quarantine made explicit; §VI.D prose fix separates fresh-IsolationForest injection-recovery cross-check from BigAE-latent ranking that drives the eROSITA + Gaia primary catalogs (closes Wave 10 cross-model BLOCKERs P3-CM-B1 + P3-OA-B2 + P3-CM-M3 + P3-OA-B6). Recompiled on Pod 3 via texlive Ubuntu (pdflatex×2 inline thebibliography): 28.27 MB PDF, 0 undef refs. Carries forward Path-C 12/12 + Wave 5 LAMOST FAIL relabel + Wave 5 B10 100k OOD validation (median MSE 0.178).	100 %	`pipelines/p3_anomaly_engine/paper3_draft.tex`
4	Galaxy Chirality Catalog	v1.0.6, 16 pp. R42 Wave 11 LANDED 2026-05-01 07:30 PDT — N_spiral=3,321,795 shot-noise correction landed in NaMaster footnote 5 (cross-confirmed: Gemini 3.1-Pro P4-CM-B1 linear-algebra MASTER deconvolution critique and GPT-5 P4-OA-M7 footnote-5 N_gal-vs-N_spiral arithmetic critique independently flagged the same arithmetic bug — `C_noise^(1)` was being computed from N_gal=8,474,531 although the asymmetry field uses only 3,321,795 spirals). Shot-noise floor up 2.55×, propagates to 2.65× C_ℓ error budget uplift; the headline equivariance result (raw +2.05% → eq −0.53%, 3.86× suppression factor) survives because it is a within-spiral-pool measurement that never used N_gal. Hemisphere look-elsewhere null v3 running on Pod 3 in parallel (max\\|A\\|=8.53e-3 at RA=78.75° Dec=−66.44°, 500 MC) to retire P4-CM-B2 PSF cross-correlation skeptic-channel. Recompiled on Pod 3 via texlive Ubuntu (pdflatex×2 inline thebibliography): 25.78 MB PDF, 0 undef refs. 8.47 M galaxies · 8/8 bias tests · 0.43σ null dipole · Shamir refuted 7×.	100 %	`pipelines/p2_chirality/chirality_catalog_paper.tex`

Program-level arXiv ETA: All four papers at 100 % on content + compile axes as of 2026-04-22. Papers 1, 2, 4 at 100 % since 2026-04-17 (Docker TeX Live recompile, 945 KB / 632 KB / 25.7 MB, 0 undef refs each). Paper 3 Path-C 12/12 CLOSED fire #189 — criterion #4 DESI 5-fold OOS k-fold executed live on A100 pod (47K SPARCL spectra, checksum 1812395110 matches dry-run, mean pairwise &Jbar; = 0.862 vs required 0.70, consensus ≥3/5 = 464/546 = 85.0 % of top-1 % union); Paper 3 Table 1 headline (195,829 / 0.87 %) stands unchanged; 28 MB / 33 pp / 0 undef PDF recompiled. Cron 9f44c29e self-terminated (weighted sum 12.000/12 = 100.000 %). Remaining items are all Houston-owned: (a) one-line review of the LSST projection line in Paper 4 (non-blocking) and (b) arXiv form submission. Recommended submission order: Paper 4 → Paper 3 → Paper 1 → Paper 2.

Cross-paper dependencies

Paper 1 (Spin-Torsion) ──┬─> theoretical f_NL = −35/8 ─> Paper 2 (Fisher forecast) │ │ └─> 14 structural barriers ├─> multi-tracer bias α ─> uses Paper 3 anomalies │ Paper 3 (Anomaly Catalog) ─┬─> AI-selected tracers ─────┘ │ └─> Shares dipole infrastructure (Landy-Szalay w(θ)) ─> Paper 4 dipole code Paper 4 (Chirality Catalog) ─> dipole/TTA code ─> available for Paper 3 limitation G

Per-paper SSOT

Spin-Torsion Cosmology from ECH Gravity

canonical: arxiv/main.tex · v2.3.5 · 1208 lines · 1.23 MB PDF (R42 Wave 11 LANDED 2026-05-01 07:30 PDT) · gr-qc primary

100%

arXiv-ready

Most mature paper in the program. 10+ revision rounds complete. ALP birefringence β = 0.242° ± 0.061° (3.9σ inverse-variance combined) already integrated. Compiles clean with 0 undefined refs. Remaining gap is polish + site sync, not science.

Verified headline claims

Claim	Value
ΔN_eff (full-tension dataset, 176,240 samples)	−0.020 ± 0.169
H₀ (standard ΛCDM)	67.68
ALP birefringence β (inverse-variance combined)	0.242° ± 0.061° (3.9σ)
Bayes factor (spectator vs trivial)	176
N_tot viable fraction [79, 95]	2.2 %
Structural barriers closed	14
Branch V matter bounce f_NL	−35/8 = −4.375
NANOGrav 15yr γ (bounce prediction 3.0)	3.20 ± 0.42 (0.48σ)

Close-the-gap queue (7 tasks · ~1 % total)

P1-LINE-299-WORDSMITH [x]✓ Done 2026-04-17: TBD rewritten to cite sec:futuredirections explicitly and reference the spectator-ALP photon-torsion coupling channel for the isotropic angle.0.2 %
P1-FIGURES-VERIFY [x]✓ Done 2026-04-17: both \includegraphics calls resolve (figure1_lqg_holst_derivation_enhanced.png, consistency_window_birefringence.pdf). PDF ≥ 1 MB check deferred to pod recompile.0.1 %
P1-CORNER-PLOTSGenerate corner plots from existing 425 k MCMC samples (getdist), embed in §IV, drop L882 "companion data release" note0.2 %
P1-PDF-RECOMPILERecompile on-pod with \paperTimestamp refreshed0.2 %
P1-SITE-SYNCSync index.html, paper.html, explained.html, activity.html, figures.html, glossary.html with v2.3.x final numbers0.1 %
P1-WIKI-SYNC [x]✓ Done 2026-04-17: paper-1-spin-torsion.md rewritten as pointer-only; stale "80 % submission-ready / TIER-1 edits" claim removed; SSOT + science-highlights links added.0.05 %
P1-TARBALL [~]Tarball built 2026-04-17 at arxiv/main_arxiv_submission.tar.gz (2.0 MB, 14 figures + main.tex + references.bib + main.bbl). Clean-revtex smoke-test from tarball alone still pending pod.0.15 %

Principle-10 audit (8 future-work hits, classified)

DO-NOW (1): L882 corner plots — existing 425 k MCMC samples, trivial via getdist → closes "companion data release" promise
WORDSMITH (1): L299 "TBD" — 15-min prose edit
TRULY BLOCKED (3): L291 photon-torsion coupling; L744 Cuscuton-ECH mapping; L976 γ-origin derivation — all require new theoretical frameworks beyond paper's scope
BENIGN (3): L306, L736, L1009 are forward-looking statements, not deferred work

Full SSOT: project-context/SSOT/paper-1/status.md
Science highlights (N0–N4): paper1_science_highlights.md — 9 contributions, N3×5 / N2×4

f_NL Forecast — SPHEREx / MegaMapper

canonical: public/papers/paper2_fnl_forecast.pdf · v1.6.1 · revtex4-2 · arXiv-ready

100%

overall / 100 % arXiv format

STATUS. Revtex4-2 conversion landed fire #9 (2026-04-17). Citations resolved, PDF compiled on pod, posted to public/papers/paper2_fnl_forecast.pdf. Peer-review follow-ups (cite Paper 3 internally, Fisher-number reconcile) tracked in queue but do not block submission.

Verified science (complete)

Claim	Value
σ(f_NL) baseline (Fisher ideal)	0.36
σ(f_NL) conservative (Munchmeyer+2019)	0.93
SPHEREx detection significance of f_NL = −4.375	4.7 – 12σ by 2027
Multi-tracer 5-tracer σ(f_NL) (anomaly-optimised)	11.71
Bayes factors (AI vs standard)	8 – 17
Gold+Silver QSO clustering bias validation	1.58× enhanced

Close-the-gap queue (9 tasks · ~15 % total)

P2-REVTEX4-2-CONVERTRewrite from article+natbib to revtex4-2 PRD + embedded \bibitem6 %
P2-BIB-RESOLVEConvert \citep{}→\cite{}, embed bibitems; zero [?] in PDF4 %
P2-COMPILE-PODRecompile on pod after conversion; verify ≥ 2 MB, 0 undefined refs, all 6 figures embedded2 %
P2-XREF-AUDIT [x]✓ Done 2026-04-17: multi-tracer language is purely SPHEREx/MegaMapper as-designed; no implicit Paper 3 reference; no cross-cite needed. Paper 1 \citep{Golden:2026framework} already in place.1 %
P2-SITE-SYNCUpdate index.html σ(f_NL) card, paper.html 85%→100%, activity.html, figures.html (+6), data-explorer.html Fisher JSON1 %
P2-WIKI-POINTER [x]✓ Done 2026-04-17: rewritten as pointer-only stub; stale v1.3.0 + "SUBMISSION-READY" claim removed; SSOT + science-highlights links added.0.3 %
P2-CURRENT-STATUS-SYNC [x]✓ Done 2026-04-17: Paper 2 row now reads "v1.6.0 · 85 % — science done, NOT arXiv-ready" with revtex4-2 blocker + SSOT link.0.2 %
P2-PDF-PUBLISHCopy compiled revtex4-2 PDF to public/papers/paper2_fnl_forecast.pdf, link from paper.html0.3 %
P2-TARBALLAssemble arXiv tarball (tex + bbl + 6 figs + bphi.pdf), smoke-test clean revtex build0.2 %

Principle-10 audit (3 hits, all TRULY BLOCKED)

SPHEREx launch — 2028 data release. Cannot fold in now.
MegaMapper — 2032+ first light. Cannot fold in now.
Scope-limit boilerplate on extended-tracer subsamples — explicitly out of scope per paper framing.

Zero DO-NOW items. Every future-work-flagged item is a real future-survey dependency, not deferred analysis.

Full SSOT: project-context/SSOT/paper-2/status.md
Science highlights (N0–N4): paper2_science_highlights.md — 7 contributions, N3×3 / N2×4

Multi-Survey Anomaly Catalog

canonical: pipelines/p3_anomaly_engine/paper3_draft.tex · Path-C rebuild 12/12 CLOSED (fire #189) · PDF 28 MB / 33 pp / 0 undef · criterion #4 DESI k-fold Jaccard gate PASS

100%

Path C CLOSED (fire #189, 12/12, 100 %)

🟢 Path-C rebuild 12/12 CLOSED (fire #189, 2026-04-22). Houston pushed back 2026-04-19 on the "58.8 % novel" headline and the cross-transfer methodology (DESI-trained BigAE applied to SDSS/LAMOST; catastrophically undertrained CMB AE). Path-C rebuild replaced cross-transfer scans with native-retrained per-survey models, added NEOWISE ecliptic mask, 8-way positional dedup at 5″, and injection-recovery on every retained survey. All 12 criteria CLOSED (#1 SDSS native re-score fire #164 / #2 LAMOST native re-score / #3 CMB native retrain 100 % injection-recovery / #4 DESI 5-fold OOS k-fold fire #189 — live SPARCL 47K spectra / checksum 1812395110 / mean pairwise &Jbar; = 0.862 vs required 0.70 / consensus ≥3/5 = 464/546 = 85.0 % of union / #5 NEOWISE ecliptic mask / #6 injection-recovery all-surveys / #7 8-way dedup 378,480 unique + 637 multi-survey clusters fire #164 / #8 Paper 3 integration fire #167 / #9 PDF recompile 28 MB / 33 pp / 0 undef / #10 HF rebuild 5-file 15.2 MB fire #166 / #11 P1-PDF-V3 / #12 site-sync). Weighted sum 12.000/12 = 100.000 %. Cron 9f44c29e self-terminated. Full per-fire tracking in SSOT/drive-to-100.md Loop log.

Pre-Path-C locked state (preserved as §7 before-after baseline). 8 surveys, 37.3 M sources, 319,443 sum-over-surveys anomalies, 58.8 % SIMBAD-unmatched at 5″. Native-retrain supersession landed fire #164: 7-way dedup at 5″ (ACT DR6 quarantined, excluded from headline) yields 378,280 unique physical objects + 637 multi-survey clusters; SDSS native re-score 1,925,279 scored (~6500× anomaly-rate reduction vs cross-transfer).

Per-survey catalog (every anomaly count traced)

Survey	Scored → Anomalies
DESI DR1	22.5 M → 195,829 (0.87 %)
SDSS DR18	2.30 M → 77,905 (3.38 %)
LAMOST DR10	11.4 M → 44,075 (0.39 %)
eROSITA DR1	930 K → 298 (68 % novel)
Planck CMB × ACT DR6	40 K patches → null cross-corr
Gaia DR3	50 K → 500 (27 % novel)
NEOWISE	43.5 K → 436 (45 % novel)
Total (Path-C post-native-retrain)	37.3 M → 319,443 · 58.8 % novel — sum-over-surveys; 7-way dedup at 5″ (ACT DR6 quarantined, excluded from headline) yields 378,280 unique physical objects + 637 multi-survey clusters (fire #164, supersedes pre-Path-C 191,632). SDSS native re-score landed fire #164: 1,925,279 scored, ~6500× anomaly-rate reduction vs cross-transfer.

Verified science-headline claims

Claim	Value
σ(f_NL) DESI alone (AI-enhanced)	+6.1 % over standard
σ(f_NL) DESI + SDSS	+16.4 %
σ(f_NL) latent-space 5-tracer	+9.5 %
SPHEREx f_NL = −4.375 detection	4.38σ
NANOGrav γ (bounce prediction 3.0)	3.20 ± 0.42 (0.48σ)
ΔBIC(SMBHB − bounce)	7.0 (strong bounce)
MCMC samples	192,000 (32×6,000)
LAMOST blue-excess artifact rate	98 % (negative control)
Unsupervised photo-z σ_NMAD	0.028

Close-the-gap queue (8 tasks · ~1 % total)

P3-PDF-CANON ✓Done 2026-04-17 — arxiv/paper3_anomaly_catalog.tex replaced with pointer stub, stale .pdf removed. Canonical is pipelines/p3_anomaly_engine/paper3_draft.tex0.3 %
P3-PDF-RECOMPILERecompile on-pod with today's date + SSOT cross-check0.3 %
P3-ATIC 374313355 periodicity analysis (TESS archival + Lomb-Scargle)0.05 %
P3-BDeep cross-match of top-100 DESI + 203 eROSITA + BAL-QSO against NED / VizieR / Gaia-XP0.1 %
P3-CFisher σ(γ) forecast for NANOGrav 20yr / EPTA DR3 / SKA-P10.05 %
P3-SITE-SYNCUpdate index.html stat cards, paper.html readiness, activity.html, figures.html, data-explorer.html with SSOT numbers0.1 %
P3-HF-UPLOADPublish 319,443-anomaly catalog to HuggingFace bamfai/bigbounce-anomaly-catalog CC-BY-4.00.05 %
P3-XREFAudit Paper 3 cross-refs against Paper 2 f_NL forecast + Paper 4 dipole infrastructure0.05 %

Principle-10 audit (4 future-work-adjacent hits — all DO-NOW or SIMULATE-NOW)

A · DO NOW: L423 TIC 374313355 follow-up → use existing ZTF periodicity code
B · DO NOW (partial): L633 follow-up spectroscopy → NED + VizieR + Gaia-XP + archival xmatch (probably shrinks novel count 20–40 %)
C · SIMULATE/AUGMENT NOW: L558 "continued monitoring" → Fisher σ(γ) shrinkage for NANOGrav 20yr / EPTA DR3 / SKA-P1
D–H · §7.3 Limitations (5 items): ensemble detection, injection-recovery, DESI B-dominant audit, empirical bias α calibration, NANOGrav raw-TOA reforecast — all DO-NOW or SIMULATE-NOW; none truly blocked

Policy: Submit v1 after A, B, C (7 days on-pod). D–H feed v2 / reviewer-response.

Full SSOT: project-context/SSOT/paper-3/status.md
Science highlights (N0–N4): paper3_science_highlights.md — 10 contributions, N3×4 / N2×6

Galaxy Chirality Catalog

canonical: pipelines/p2_chirality/chirality_catalog_paper.tex · 1,099 lines · 19 MB PDF · 11 figures

97%

1–2 days to submit

Science DONE. 8,474,531 galaxies classified · bias tests 8/8 pass · dipole is a null (0.43σ, p=0.33) · Shamir 2020 3 % asymmetry claim refuted 7×. Dipole JSON gap closed 2026-04-17. arXiv submission blocked on 4 trivial admin items (~45 min), not science.

Verified headline claims

Claim	Value
Galaxies classified	8,474,531
Classification accuracy (3-class)	93.7 %
Bias tests passed	8 / 8
Raw (A) dipole — spurious artifact	94.6σ
Equivariant (C) dipole — null	0.43σ (p = 0.33)
CW/(CW+CCW) equivariant	0.4974
CE-ResNet external agreement	91.5 % on 23 k galaxies
Max regional asymmetry	0.47 %
Shamir (2020) claim	~3 % — refuted 7×

Close-the-gap queue (7 tasks · ~3 % total)

P4-DIPOLE-JSON-REBUILDRe-emit non-truncated outputs/dipole/summary.json (current one crashed mid-write at L366)0.5 %
P4-PDF-CANON ✓Done 2026-04-17 — arxiv/paper4_chirality_catalog.tex replaced with pointer stub, stale .pdf removed. Canonical is pipelines/p2_chirality/chirality_catalog_paper.tex0.5 %
P4-PDF-RECOMPILERecompile on-pod with today's date + SSOT cross-check0.5 %
P4-PAPER2-XREF ✓Done 2026-04-17 — L234–238 rewritten past-tense (full catalog + 8/8 audit); L255 replaced with companion-paper cross-ref citing 0.43σ null + 7× Shamir refutation0.3 %
P4-SITE-SYNC ~Partial 2026-04-17 — index.html + paper.html done (85%→97%, stale links → canonical, highlights links). activity.html, figures.html, data-explorer.html still pending.0.3 %
P4-HF-DOIPin HF bamfai/galaxy-chirality-catalog version + add DOI / versioned URL to data-availability0.2 %
P4-LSST-LINE-REVIEWHouston reviews L913 "Future surveys (Rubin LSST)" line on final PDF — confirm TRULY BLOCKED0.2 %

Principle-10 audit (1 TRULY BLOCKED, 2 OUTDATED-WORDING-FIX, 0 DO-NOW)

TRULY BLOCKED: L913 "Future surveys" (Rubin LSST Y1 not yet public) — legitimate future-survey statement, keep as-is
OUTDATED WORDING: paper2_chirality_section.tex:237 "pending at the time of writing" → rewrite past tense (catalog is complete)
OUTDATED WORDING: paper2_chirality_section.tex:255 "will be presented in future work" → rewrite as Paper-4 cross-reference
Earlier false alarms resolved: CE-ResNet cross-val DONE (91.5 %); dipole DONE (git 5d24cfc); redshift-binned f_cw DONE (20 bins)

Zero DO-NOW blockers. Paper 4 is honestly ready.

Full SSOT: project-context/SSOT/paper-4/status.md
Science highlights (N0–N4): paper4_science_highlights.md — 7 contributions, N3×3 / N2×3 / N1×1

Cross-paper task queue

Prioritized queue across all four papers + program-level tasks. Filter by priority, paper, or owner. Full queue at project-context/SSOT/queue.md.

Pri	ID	Title / notes	Owner	Paper	%	Status
P0	P2-REVTEX4-2-CONVERT	Rewrite Paper 2 .tex from article+natbib to revtex4-2 PRD Single biggest blocker to Paper 2 submission	agent	P2	6 %	[ ]
P0	P2-BIB-RESOLVE	Convert \citep→\cite, embed bibitems; zero [?] in PDF Follows P2-REVTEX4-2-CONVERT	agent	P2	4 %	[ ]
P0	P3-A	TIC 374313355 periodicity analysis TESS archival light curve + Lomb-Scargle; uses ZTF-anomaly periodicity code	pod	P3	0.05 %	[ ]
P0	P3-B	Deep cross-match of top-100 DESI + 203 eROSITA + BAL-QSO NED / VizieR / Gaia-XP; probably shrinks novel count 20–40%	agent	P3	0.1 %	[ ]
P0	P4-DIPOLE-JSON-REBUILD	Re-emit non-truncated outputs/dipole/summary.json Current one crashed at L366 after consistent_with_null	pod	P4	0.5 %	[ ]
P1	P1-LINE-299-WORDSMITH	Replace (amplitude and shape TBD) on L299 of arxiv/main.tex Last TBD in Paper 1; 15-min edit	agent	P1	0.2 %	[x]
P1	P1-FIGURES-VERIFY	Verify every \includegraphics in arxiv/main.tex resolves PDF is currently only 510 KB (low) — pre-recompile disk check	agent	P1	0.1 %	[x]
P1	P1-CORNER-PLOTS	Generate corner plots from existing 425k MCMC samples getdist, embed in §IV, drop L882 "companion data release" note	pod	P1	0.2 %	[ ]
P1	P1-PDF-RECOMPILE	Recompile Paper 1 PDF on-pod with \paperTimestamp refreshed After wordsmith + corner plots	pod	P1	0.2 %	[ ]
P1	P1-SITE-SYNC	Sync index/paper/explained/activity/figures/glossary pages Part of P-SITE-FULL-SYNC	site	P1	0.1 %	[ ]
P1	P1-WIKI-SYNC	Freeze wiki/entities/paper-1-*.md as pointer-only files Mechanical	agent	P1	0.05 %	[x]
P1	P1-TARBALL	Build Paper 1 arXiv tarball + smoke-test clean revtex build Tarball built 2026-04-17; pod smoke-test pending	agent	P1	0.15 %	[~]
P1	P2-COMPILE-POD	Recompile Paper 2 after revtex4-2 conversion Verify ≥2 MB, 0 undefined refs, all 6 figures embedded	pod	P2	2 %	[ ]
P1	P2-XREF-AUDIT	Audit Paper 2 cross-refs against Paper 1 framework + Paper 3	agent	P2	1 %	[x]
P1	P2-SITE-SYNC	Update σ(f_NL) card, paper.html 85→100, figures + Fisher JSON	site	P2	1 %	[ ]
P1	P2-WIKI-POINTER	Rewrite wiki/entities/paper-2-fnl-forecast.md as pointer Current file stale 2026-04-04, wrongly claims SUBMISSION-READY	agent	P2	0.3 %	[x]
P1	P2-CURRENT-STATUS-SYNC	Update Paper 2 row in CURRENT_STATUS.md "v1.3.0 · Ready" → "v1.6.0 · 85 % · revtex4-2 conversion pending"	agent	P2	0.2 %	[x]
P1	P2-PDF-PUBLISH	Copy compiled revtex4-2 PDF to public/papers/, link from paper.html After P2-COMPILE-POD	pod	P2	0.3 %	[ ]
P1	P2-TARBALL	Assemble Paper 2 arXiv tarball (tex + bbl + 6 figs + bphi.pdf)	agent	P2	0.2 %	[ ]
P1	P3-C	Fisher σ(γ) for NANOGrav 20yr / EPTA DR3 / SKA-P1 Addresses §6 "continued monitoring" with concrete when-decisive figure	agent	P3	0.05 %	[ ]
P1	P3-PDF-CANON	Delete or rebuild arxiv/paper3_anomaly_catalog.tex + .pdf Done 2026-04-17: .tex replaced with pointer stub, .pdf removed	agent	P3	0.3 %	[x]
P1	P3-PDF-RECOMPILE	Recompile Paper 3 PDF on-pod with today's date + SSOT cross-check ~15 min on H200	pod	P3	0.3 %	[ ]
P1	P4-PDF-CANON	Pick pipelines/p2_chirality/chirality_catalog_paper.tex canonical Done 2026-04-17: arxiv/.tex replaced with pointer stub, .pdf removed	agent	P4	0.5 %	[x]
P1	P4-PDF-RECOMPILE	Recompile Paper 4 PDF on-pod with today's date + SSOT cross-check	pod	P4	0.5 %	[ ]
P1	P4-PAPER2-XREF	Fix 2 stale wordings in paper2_chirality_section.tex Done 2026-04-17: L234-238 past-tense rewrite + L255 Paper-4 cross-reference with null 0.43sigma cite	agent	P4+P2	0.3 %	[x]
P1	P3-SITE-SYNC	Update index / paper / activity / figures / data-explorer to reflect SSOT	site	P3	0.1 %	[~]
P1	P4-SITE-SYNC	Site sync for Paper 4 SSOT numbers index.html + paper.html updated 2026-04-17 (85%→97% ready, stale arxiv .tex/.pdf links replaced with canonical, science-highlights links added). activity.html / figures.html / data-explorer.html still pending.	site	P4	0.3 %	[~]
P1	P3-HF-UPLOAD	Publish 319,443-anomaly catalog to HuggingFace CC-BY-4.0 Paper §9 data-availability needs live link before arXiv	agent	P3	0.05 %	[ ]
P1	P4-HF-DOI	Pin HF bamfai/galaxy-chirality-catalog version + add DOI	agent	P4	0.2 %	[ ]
P1	P3-XREF	Audit Paper 3 cross-refs against Paper 2 f_NL + Paper 4 dipole	agent	P3	0.05 %	[ ]
P1	P4-LSST-LINE-REVIEW	Houston reviews L913 "Future surveys (Rubin LSST)" Confirm TRULY BLOCKED; if not, becomes new task	Houston	P4	0.2 %	[ ]
P1	P-MEMORY-SYNC	Add MEMORY.md entry for SSOT dir + update entries referencing old paths	agent	ALL	0 %	[ ]
P2	P3-D	Ensemble anomaly detection (VAE + iForest + one-class SVM) §7.3 #1. 2–3 wk on H200	pod	P3	0.05 %	[ ]
P2	P3-E	Synthetic-anomaly injection + recovery for 7 non-DESI surveys §7.3 #2. ~2 wk each, parallelisable	pod	P3	0.1 %	[ ]
P2	P3-F	DESI B-dominant population (44,436 / 22.7%) calibration audit §7.3 #3. ~2 wk	pod	P3	0.05 %	[ ]
P2	P3-G	Empirical Landy-Szalay w(θ) bias calibration for anomaly subsample §7.3 #4. Replaces α = 0.15 assumption; re-uses Paper 4 dipole infrastructure	pod	P3+P2	0.1 %	[ ]
P2	P3-H	NANOGrav reforecast with inflated DR3 free-spectrum covariance §7.3 #5. ~1 wk	agent	P3	0.05 %	[ ]
P2	P-SITE-FULL-SYNC	Site-agent pass: run all P{1,2,3,4}-SITE-SYNC variants together See index.md aggregate surface-sync checklist	site	ALL	~1 %	[ ]
P2	P-ARXIV-P4	Assemble Paper 4 tarball, fill arXiv form, submit, return ID Do Paper 4 first (most self-contained)	Houston	P4	closes	[ ]
P2	P-ARXIV-P3	Submit Paper 3 to arXiv Follow ~24 h after Paper 4	Houston	P3	closes	[ ]
P3	P-MEMORY-AGENT-HOOKS	Ensure AGENTS.md + CLAUDE.md route every agent to SSOT first Partially done in restructure commit; verify after 1 session	agent	ALL	0 %	[ ]
P3	P-LEGACY-STATUS-CLEAN	Rewrite CURRENT_STATUS.md as mirror of SSOT/index.md	agent	ALL	0 %	[ ]
P3	P-FREEZE-WIKI	Confirm all wiki/entities/paper-.md and pipeline-.md are pointer-only	agent	ALL	0 %	[ ]
P3	P-SSOT-CRON	Weekly cron/agent check: SSOT file > 7 days stale → flag Prevents drift	agent	ALL	0 %	[ ]

Completed this session (2026-04-17)

P4-DIPOLE-ARTIFACTS-LOCAL — Copied dipole summary + figures + log from pod_final_backup_20260414/ to pipelines/p2_chirality/outputs/dipole/ · commit 6651dd5
P4-REDSHIFT-BINS-INDEX — Indexed the already-done fcw_vs_redshift.csv (20 bins) · part of 6651dd5
P3-SSOT-SWEEP — Forensic sweep on Paper 3 · produced SSOT/paper-3/status.md · commit 0c39a15
P4-SSOT-SWEEP — Forensic sweep on Paper 4 · produced SSOT/paper-4/status.md
P-SSOT-RESTRUCTURE — Moved to SSOT/ tree; added README, index, queue; pointer stubs kept at old paths · commit ae21ac5
P-PRINCIPLE-10-CORRECTION-P3 — Corrected Paper 3 "zero future-work hits" claim after Houston pushback; 4 hits now classified DO-NOW / SIMULATE-AUGMENT-NOW · commit ae21ac5
P1-SSOT-SWEEP — Forensic sweep on Paper 1 · produced SSOT/paper-1/status.md at 99 %
P2-SSOT-SWEEP — Forensic sweep on Paper 2 · produced SSOT/paper-2/status.md at 85 % with revtex4-2 blocker
P-SSOT-VIEWER — Built ssot.html viewer page on bigbounce site (this page) · wiring nav entry in nav.js
P-SCIENCE-HIGHLIGHTS — Wrote/updated paper{1,2,3,4}_science_highlights.md with N0–N4 novelty tags. Paper 1: N3×5/N2×4. Paper 2: N3×3/N2×4. Paper 3: N3×4/N2×6 (expanded from DESI-only to full 8-survey). Paper 4: N3×3/N2×3/N1×1. Cross-linked from each SSOT/paper-N/status.md and from each panel on this page · commit 47413e1
P3-PDF-CANON — Replaced arxiv/paper3_anomaly_catalog.tex with a pointer stub and removed the stale .pdf. Canonical is pipelines/p3_anomaly_engine/paper3_draft.tex
P4-PDF-CANON — Replaced arxiv/paper4_chirality_catalog.tex with a pointer stub and removed the stale .pdf. Canonical is pipelines/p2_chirality/chirality_catalog_paper.tex
P4-PAPER2-XREF — Rewrote paper2_chirality_section.tex L234–238 past-tense (full 8/8 audit + complete catalog) and L255 with companion-paper cross-ref citing the 0.43σ null and the 7× Shamir refutation
P4-SITE-SYNC (partial) — index.html and paper.html updated: 85%→97% Ready, stale arxiv/paper4_* links replaced with canonical pipelines/p2_chirality/ paths, science-highlights links added. activity.html, figures.html, data-explorer.html still pending.
P3-SITE-SYNC (partial) — index.html Paper 3 card + stat cards now quote SSOT canonical 319,443 anomalies / 37.3M sources / 58.8% SIMBAD-unmatched; paper.html subtitle, “How these papers fit together”, Paper 3 listing + title + key-results bullet all updated; activity.html received a 2026-04-17 timeline entry covering the SSOT-science-highlights rollout and P3/P4 task closures. figures.html and data-explorer.html still pending.
P2-CURRENT-STATUS-SYNC — project-context/CURRENT_STATUS.md Paper 2 row: "v1.3.0 · Ready for submission" → "v1.6.0 · 85 % — science done, NOT arXiv-ready" with revtex4-2 blocker + SSOT link.
P2-WIKI-POINTER — wiki/entities/paper-2-fnl-forecast.md rewritten as pointer-only; stale 2026-04-04 "SUBMISSION-READY" claim removed; SSOT + science-highlights links added.
P2-XREF-AUDIT — grep of 02_full_draft.tex confirmed multi-tracer language is purely SPHEREx/MegaMapper as-designed (Heinrich 2023, Schlegel 2022); no implicit Paper 3 reference; no cross-cite needed. Paper 1 handle \citep{Golden:2026framework} already in place.
P1-LINE-299-WORDSMITH — arxiv/main.tex L299 "amplitude and shape TBD" replaced with explicit “not derived in this work; flagged as open in §futuredirections” phrasing. No new science claimed.
P1-FIGURES-VERIFY — both \includegraphics paths in arxiv/main.tex resolve on-disk (figure1_lqg_holst_derivation_enhanced.png, consistency_window_birefringence.pdf). PDF size check (≥1 MB) deferred to post-recompile.
P1-WIKI-SYNC — wiki/entities/paper-1-spin-torsion.md rewritten as pointer-only; stale "~80 % submission-ready / four TIER-1 edits" removed; SSOT + science-highlights links added.
P1-TARBALL (partial) — arxiv/main_arxiv_submission.tar.gz built (2.0 MB, 14 figures + main.tex + references.bib + main.bbl). Clean-revtex smoke-test from the tarball alone still pending pod (requires texlive-publishers).

How to work this queue

Pick the lowest-ID P0 task whose dependencies are met.
Flip status to [~] in project-context/SSOT/queue.md when you start.
When done, flip to [x] and move the row to the Completed section with the commit hash.
If blocked, flip to [!] with a one-line reason.
When adding new tasks: insert at correct priority; ID them P{paper}-{slug} or P-{program-slug}.
Never delete a completed task row. That's the record.
Update this viewer (ssot.html) in the same commit as any status / queue change so the site doesn't drift from the canonical SSOT.

SSOT & Task Queue

Program health at a glance

Cross-paper dependencies

Per-paper SSOT

Spin-Torsion Cosmology from ECH Gravity

fNL Forecast — SPHEREx / MegaMapper

Multi-Survey Anomaly Catalog

Galaxy Chirality Catalog

Cross-paper task queue

Completed this session (2026-04-17)

How to work this queue

f_NL Forecast — SPHEREx / MegaMapper