diff options
| author | Yuren Hao <yurenh2@illinois.edu> | 2026-07-16 04:01:03 -0500 |
|---|---|---|
| committer | Yuren Hao <yurenh2@illinois.edu> | 2026-07-16 04:01:03 -0500 |
| commit | 5f8dd1c4d7106665cf9722e7a394893bb87c9b07 (patch) | |
| tree | 021cb33e97e5454bf866205c739da9fe7fe1a981 | |
| parent | 4b58cf8ba2e8e9ba167dc61dc4ed2e58dfdab4d8 (diff) | |
RESULT 28 caveat: fp32-EP control insufficient — BP mirror is the decision metric (user correction)
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_014FAPDWQ49M5Ye3NpTndTpn
| -rw-r--r-- | docs/campaign/CASCADE_ABLATION_PLAN.md | 6 |
1 files changed, 6 insertions, 0 deletions
diff --git a/docs/campaign/CASCADE_ABLATION_PLAN.md b/docs/campaign/CASCADE_ABLATION_PLAN.md index 7542863..12751d9 100644 --- a/docs/campaign/CASCADE_ABLATION_PLAN.md +++ b/docs/campaign/CASCADE_ABLATION_PLAN.md @@ -586,6 +586,12 @@ Monotone dose-response; clean (zero guards). Readings: qcomp8/6/4 arms running (gate #1b). - Procurement impact: the Y3/pc AD7528 line is unaffected either way (its role is compute); the acceptance ENOB bar keys on the qcomp verdict. +- CAVEAT (user, 2026-07-16): the fp32-EP control conflates "quantization tax" with "EP-specific + quantization tax". BP MIRROR SUITE chained (bp_qctl + qup10/8/6 + qcomp8/6/4 on the BP twin, + same resume protocol). Decision metric = Delta_EP(bits) vs Delta_BP(bits): equal deltas => + generic quantized-training problem => the BP QAT/shadow toolbox transfers to EP unchanged. + Reading-one above ("naked cells need >=10-12b") is PROVISIONAL until the BP control lands — + naked quantized writes likely hurt BP comparably. ### RESULT 27 (2026-07-15): K-SATURATION ACROSS TRAINING (Alexi's challenge answered with data). Challenge (Alexi Gladstone): K=3 nudge sweeps seems very few; as training roughens the landscape, |