Skip to content

Record: SP8192 + Improved Parallel Residuals + Muon 0.97 + TTT 5ep + N-gram Tilt + Hessian SDClip — val_bpb 1.07730#1557

Open
ndokutovich wants to merge 1 commit intoopenai:mainfrom
ndokutovich:submission-s8-ha
Open

Record: SP8192 + Improved Parallel Residuals + Muon 0.97 + TTT 5ep + N-gram Tilt + Hessian SDClip — val_bpb 1.07730#1557
ndokutovich wants to merge 1 commit intoopenai:mainfrom
ndokutovich:submission-s8-ha

Conversation

@ndokutovich
Copy link
Copy Markdown

Record: SP8192 + Improved Parallel Residuals + Score-First TTT + Causal N-gram Tilt + Hessian SDClip

val_bpb = 1.07730 (3-seed mean, std 0.00040) | ~15.97 MB | 8xH100 SXM

3-Seed Results

Seed Sliding BPB TTT BPB Artifact
42 1.07880 1.07684 15,965,495
314 1.07959 1.07748 15,965,495
999 1.07963 1.07757 15,965,495
Mean 1.07934 1.07730

Merged SOTA (PR #1493): 1.0810. Delta: -0.00370 nats.

Techniques

  • Architecture: SP8192, 11L x 512d, 8H/4KV, MLP 4x, improved parallel residuals (L7+)
  • Training: Muon 0.97 (row-normalized), Matrix LR 0.03, EMA 0.997, 3-layer depth recurrence (L3-5)
  • Quantization: GPTQ int6 (attn+MLP) + int8 (embeddings) + Hessian-Aware SDClip (lambda=0.175)
  • Eval: Score-first TTT (SGD, 5 epochs, lr=0.005) + Causal n-gram tilt (beta=2.0, agree=0.1)
  • Compression: Brotli quality=11

Compliance

Attribution

Compute

Funded by OpenAI Advanced Competitor grant ($500 RunPod credit). 8xH100-SXM, ~3 runs for 3 seeds.

…t TTT 5ep + Causal N-gram Tilt + Hessian SDClip — val_bpb 1.07730 (3-seed mean)
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant