WeightBench - TR-04 Write: Zero-Shot Weight Synthesis (6): leaderboard
Default difficulty, seed0. Claude2samples/task; DeepSeek/GPT1sample/task. Quarantined/invalid submissions score0 (unreviewed lower bounds). Task scores average across runs then tasks; macro averages5tracks. GPT excludes4 loop Fix tasks; use common38task macro, never pool GPT9taskFix with13taskFix. Metric: Native S_WB (0–1). Source: https://arxiv.org/html/2611.18553v1
Metric: Native S_WB (0–1). Source: arxiv.org. Saturation forecast: Around February 2027. 5 models tracked. Updated 17 Nov 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-6.1 Sol | 0.863 | #12 |
| 2 | Claude Opus 5.5 | 0.829 | #7 |
| 3 | Claude Sonnet 5.5 | 0.569 | #21 |
| 4 | Claude Haiku 5.5 | 0.369 | #76 |
| 5 | DeepSeek V4 Flash | 0 | #265 |
No result here: #1 Claude Fable 5.2, #2 GPT-6.2 Astra, #3 Gemini 4 Carbon.
Interactive version: theaggregate.ai/benchmark?slug=weightbench-tr-04-write-zero-shot-weight-synthesis-6 · How It Works · Data refreshed daily, snapshot 2026-11-17.