N-gram logit boost + HedgeMixer + score-first TTT

[haimianbaobao007]

Summary

• N-gram logit boost: Properly normalized via softmax. Fixes hash collision normalization bug. Uses log-count boost instead of raw probability ratio.
• HedgeMixer: Online multiplicative weights mixing (neural vs neural+ngram experts).
• SGD momentum=0.95 TTT with per-layer LR and Polyak averaging.
• Online bias correction: Per-document logit bias (Nacrith 2026).
• Numba JIT acceleration, EMA skip, FA fallback chain.

Results (RTX PRO 6000, 1 GPU, 535 steps)

• FP32 base: 1.62 BPB
• Int6 + sliding window + n-gram + HedgeMixer: 2.87 BPB (-8% vs int6 alone)
• 535 steps insufficient for QAT convergence; 8xH100 expected to be much better.

Compliance

• Score-first: every token scored before any update that uses it.
• N-gram tables updated after scoring. HedgeMixer weights updated after scoring.

Based on PR #549 by @abaybektursun.

🤖 Generated with Claude Code

[immartian]

The HedgeMixer (multiplicative weights for neural vs neural+ngram experts) is a clean approach to the interpolation problem. We're working on a similar idea in PR #541 where binding energy acts as the per-token confidence signal for mixing, but your online multiplicative weights adaptation is more principled for the non-stationary eval setting.

The log-count boost for n-gram normalization is a good fix — raw probability ratios from hash tables with collisions are noisy. Have you tried scaling the boost by an IDF-like term (inverse document frequency of the n-gram context)? In our experiments, rare contexts carry much more predictive signal than common ones, and weighting by context specificity improved pattern selection significantly.

535 steps on a single GPU is tough — the n-gram + HedgeMixer results should improve dramatically with more training. Good luck on the 8xH100 run.

[MatoTeziTanka]

Community Review — N-gram logit boost + HedgeMixer + score-first TTT

BPB: 2.8700 | Compliance: LOOKS CLEAN — score-first-per-chunk TTT (legal #1413 dexhunter pattern)

What I found in the code (head SHA a6857b36bd6e, file train_gpt.py):

The TTT path at line 1340 implements the score-first-per-chunk pattern: each chunk is scored under torch.no_grad() / inference_mode() before the base_model.train() + SGD adaptation runs on that same chunk, with an is_last_chunk guard so the final chunk gets no adaptation pass. This is the structural shape of the current leaderboard's legal frontier (PR #1413 dexhunter, the 1.0828 SP8192 + QK-Gain 5 + Legal TTT entry — verified at its head SHA against the is_last_chunk + torch.no_grad() score-first accumulator pattern).

Per Issue #402 and Issue #677, TTT is legal when each token is scored before the adapter updates on it, and that's what the code does here — chunk ci is scored under weights adapted only on chunks 0..ci-1. No prequant_ttt_adapt_adamw(val_tokens, ...) multi-epoch fine-tune, no scored-region SLOT, no target-in-key n-gram cache.

CPU smoke test (CT2038 proteus-engine, 2026-04-11): import OK in 0.04s, dim=512, layers=11, vocab=1024, code=108102 B, SMOKE_TEST_PASS

Verdict: LOOKS CLEAN.

Recommendation to @cocohearts @valerio-oai @0hq @yuzhougu-oai @notapplica: MERGE pending standard checks (3-seed validation, 16MB artifact cap, 10-min wallclock on 8×H100 SXM). The compliance picture matches the legal reference frontier and no flags were raised by the classification pass.

Auto-classification caveat: this review was drafted by the AST-based classifier against a template derived from manually-reviewed cluster PRs (#1420, #1450, #1487, #1541, #1529, #1533, #1518). If I've misread a subtlety in your eval path — e.g., multi-epoch TTT that I mistook for single-pass, or a target-in-key lookup I missed in a helper function — please flag it and I'll re-run the audit manually.

---

Reviewed by @MatoTeziTanka — The Agora. CPU smoke test (CT2038 proteus-engine, 2026-04-11): import OK in 0.04s, dim=512, layers=11, vocab=1024, code=108102 B, SMOKE_TEST_PASS. Classification via deterministic AST-based classify_prs.py (pattern bank derived from ~65 manually-reviewed PRs earlier in the 2026-04-11 sweep). This review was auto-drafted from a template and spot-checked before posting — if the template misread your code, please call it out so I can iterate the classifier.