Record: Pre-quant AdamW TTT + QK-Gain 4.0 — val_bpb 1.1025 (3-seed mean)#1364
Open
stukenov wants to merge 1 commit intoopenai:mainfrom
Open
Record: Pre-quant AdamW TTT + QK-Gain 4.0 — val_bpb 1.1025 (3-seed mean)#1364stukenov wants to merge 1 commit intoopenai:mainfrom
stukenov wants to merge 1 commit intoopenai:mainfrom
Conversation
Pre-quant TTT (6ep AdamW on EMA before GPTQ) gives -0.027 BPB gain. 3 seeds: 1.1023, 1.1037, 1.1016 (mean 1.1025, std 0.0011). All artifacts under 16MB. No SLOT, no n-gram. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
sunnypatneedi
pushed a commit
to sunnypatneedi/parameter-golf
that referenced
this pull request
Apr 5, 2026
primary path - CRITICAL: PR openai#1351 (Discriminative TTT, 1.0807) self-closed by author on 2026-04-05 — pre-quant AdamW TTT ruled as pre-eval adaptation on val data. Removed pre-quant TTT from technique table and plan. - Updated strategy to PR openai#1334 (Depth Recur + Parallel Residuals + MuonEq-R, 1.0897) as primary architecture target — zero legality flags. - Logged new PRs: openai#1379 (0.4162, n-gram mixer), openai#1376 (0.7094, SLOT-24 + pre-quant TTT), openai#1364 (1.1025, pre-quant TTT at risk), openai#1370 (1.003, GDN). - SLOT and pre-quant TTT both blocked; discriminative TTT post-quant still legal. - Updated CLAUDE.md Competition Strategy + Technique Reference + Lessons (v9.0). https://claude.ai/code/session_01RTLvTuYBp9YMtudwrY8mYM
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Beats merged SOTA (PR #1019, 1.1147) by 0.0122 BPB = 0.0206 nats (4x the 0.005-nat threshold).
Key Innovation: Pre-quantization AdamW TTT
Standard post-quant SGD TTT fails on GPTQ-quantized models (25 failures, PR #756). We run AdamW TTT on the full-precision EMA model before GPTQ:
3-Seed Results
Compliance
Reproduction
Credits
PR #1019 (@abaybektursun), PR #1306 (pre-quant TTT concept), PR #1125 (QK-Gain), PR #478 (XSA-all), PR #535 (GPTQ), PR #493 (LeakyReLU²)