# Canonical numbers for the public page

**Evidence state:** 2026-07-16 · claim-stack v1 · **3/3** candidate seals  
**Resolution order:** sealed `*_SUMMARY.md` / `MULTI_SEED_CLAIM.md` → pulled CSVs → this file → page copy  
**Do not** publish % vs historical F1 (~229.1 min). **Do not** claim matched first-cross ΔT% vs joint stock.

Paths below are under `/home/q/Downloads/software_i_built/modal/` unless noted.

---

## Public denominator — F1_seal

| Field | Canonical | Source |
|---|---|---|
| Workspace | **Clock-W** | public code; real profile in `WORKSPACE_MAP.local.md` only |
| GPU | NVIDIA A100-SXM4-40GB | same |
| torch | 2.13.0+cu130 | same |
| First-cross step | 4736 | same + `nocap-repo/results/f1_seal/log.csv` |
| val @ cross | 3.381538 | same |
| `train_time_ms` @ cross | **13,918,667.5** | same |
| **T (published)** | **3.866 h** | SUMMARY (CSV exact ÷3.6e6 = 3.866297; publish **3.866**) |
| Terminal val @ 4768 | 3.379813 | same |
| Historical F1 | ~229.1 min provisional only | SUMMARY — **never** for published % |

ΔT% formula (all published %): `100 × (3.866 − T_h) / 3.866`

---

## Same-workspace check — F1_Claim-W2

| Field | Canonical | Source |
|---|---|---|
| Workspace | **Claim-W2** | public code; local file `F1_BCAROL_SUMMARY.md` |
| First-cross | @4768 · val 3.380977 | same + CSV |
| `train_time_ms` | 13,889,386.9 | same |
| **T (published)** | **3.858 h** (3.8582 detailed) | same |
| vs F1_seal | ~**+0.2%** | same |

---

## Candidate claim-stack (timing-cheap split-QKV)

Recipe pin (all three seeds): B32×accum16 · Muon LR 0.02 · mom 0.95 · Muon WD 0.1 · Adam LR 0.0018 · Adam WD 0.1 · NS 5 · zloss 0 · frozen unigram prior · raw FineWeb · wu **476** / wd **1192** · **4768** steps · A100-SXM4-40GB · torch 2.13 · `muon_split_qkv=1` · `match_joint=0`.  
Sources: `PREREG.md`, `MULTI_SEED_CLAIM.md`.

| Seed | Workspace | Cross | val @ cross | T (pub) | ΔT% vs F1_seal | Terminal CE | Stock | CSV recompute |
|---:|---|---:|---:|---:|---:|---:|---|---|
| **1337** | Claim-W1 | 4736 | 3.3741 | **3.553 h** | **+8.1%** | 3.3710 | final 3.3848 · **no cross** | no local cand CSV — sealed SUMMARY only |
| **2029** | Claim-W2 | 4736 | 3.3726 (3.372588) | **3.567 h** (3.5669) | **+7.7%** | 3.3695 (3.369498) | final 3.3840 (3.384017) · **no cross** | `…/seed2029/cand_log.csv` · ms **12,840,827.6** → 3.566897 h · ΔT% 7.74 → **+7.7%** |
| **4242** | Claim-W1 | 4736 | 3.3733 (3.373309) | **3.511 h** (3.5112) | **+9.2%** | 3.3702 (3.370249) | **not run** | `…/seed4242/cand_log.csv` · ms **12,640,470.5** → 3.511242 h · ΔT% 9.18 → **+9.2%** |

**Aggregates (from published rounded T):**

| Aggregate | Value |
|---|---|
| Mean cand T | **≈ 3.544 h** (exact mean of 3.553/3.567/3.511 = 3.543667) |
| Mean ΔT% vs F1_seal | **≈ +8.3%** (mean of +8.1/+7.7/+9.2 = +8.337…) |
| Seed 2029 vs F1_Claim-W2 | **+7.5%** |
| Bar widths vs F1 (display) | 91.9% · 92.3% · 90.8% |

All three candidates first-cross at step **4736**.

---

## Stock control (joint QKV · same stack minus split)

| Seed | Terminal CE | First-cross | Source |
|---:|---:|---|---|
| 1337 | 3.3848 | **NONE** | `SEED1337_SUMMARY.md` |
| 2029 | 3.3840 | **NONE** | `SEED2029_SUMMARY.md` + `seed2029/stock_log.csv` |
| 4242 | — | not run | `SEED4242_SUMMARY.md` |

Matched candidate-vs-stock first-cross ΔT% = **N/A**.

---

## Defensible one-liner

> On three independent seeds (1337, 2029, 4242), the claim-stack candidate (Muon B32 + timing-cheap split-QKV) first-crosses val ≤ 3.3821 about ~8% faster than our sealed stock AdamW F1 baseline (3.866 h; mean ΔT% ≈ +8.3%), with terminal CE ≈ 3.37. Where measured, the joint-QKV stock control misses the quality bar.

**Attribution:** Keller Jordan Muon + timing-cheap split-QKV (this campaign).  
**Status language:** multi-seed under-baseline vs sealed F1 — **not** an official NoCap win; pending organizer rules/submission.

---

## Seed 4242 ops note (non-metric)

First spawn rejected (A100-SXM4-**80GB**). Sealed run is relaunched **40GB**. App `ap-cl3lzJm0tSNWHMdBp0LPuX` · `runs/claim_stack_s3_cand_seed4242_4768`. Do not put secrets on the public page.