From ed5455277b8612a4635b27b770b65ff3b6f4917d Mon Sep 17 00:00:00 2001 From: Nils Date: Wed, 15 Jul 2026 21:01:23 +0200 Subject: [PATCH] pre-register item 17: GSM-only prompt-only arm (last missing cell of the GSM question) Co-Authored-By: Claude Fable 5 --- results-loop/PROTOCOL_UNIFIED.md | 12 ++++++++++++ 1 file changed, 12 insertions(+) diff --git a/results-loop/PROTOCOL_UNIFIED.md b/results-loop/PROTOCOL_UNIFIED.md index ca1bebf..2192ae4 100644 --- a/results-loop/PROTOCOL_UNIFIED.md +++ b/results-loop/PROTOCOL_UNIFIED.md @@ -324,3 +324,15 @@ h2048 single seed) were upper-tail draws of the same distribution the merge's 46.4 came from. No recipe amendment; the abstract's original "same level within noise" framing stands; single-cell records are not levels — only seed means are. + +17. **GSM-only, current recipe (pre-registered 2026-07-15 ~20:45, before + running).** train_merge_unified.py --tasks gsm: MergeAdapter, prompt- + only loop, curriculum, GSM8K data ONLY — removes the mixed-task + interference confound from the adapter_uni run, completing the + "winning recipe trained on GSM" question. Eval: prompt-only, n=256, + ks 0,1,2,4, e400. Predictions: (a) hard <= 10% at every k (supervision + density is structural: ~3 answer tokens; the recipe's dense-output + ingredient cannot exist here); (b) easy damaged at k>0 (to 40-70%); + (c) overall never beats k=0. If hard exceeds 15% or overall beats + k=0, task interference in the mixed run was masking a real GSM + capability — would reopen the GSM chapter.