PHYRD - Reproduce - Overall WAcc: leaderboard
PHYRD v1:63 Reproduce and43 Discover physics problems; complete problem-set denominators. SAcc averages each problem's correct-subquestion fraction, WAcc requires all subquestions correct. Missing/incomplete conclusions score0. Fields retain their native subsets. GPT-5.6 Sol uses a conversational interface with xhigh effort, distinct from four API models. Source-reported model identities retained without inventing dated endpoints. Reproduce. Closed-book collection; truncated responses may have additional calls, so this is not single-call pass@1. API effort is not separately documented for this collection. Overall; WAcc. Metric: Accuracy (%).
Metric: Accuracy (%). Source: arxiv.org. Saturation forecast: Around October 2027. 5 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.6 Sol (xHigh) | 68.25 | #20 (GPT-5.6 Sol) |
| 2 | DeepSeek V4 Pro | 65.08 | #109 |
| 3 | GLM-5.2 | 53.97 | #78 |
| 4 | Qwen 3.7 Max | 52.38 | #81 |
| 5 | Kimi K2.6 | 49.21 | #112 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #6 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=phyrd-reproduce-overall-wacc · How It Works · Data refreshed daily, snapshot 2026-10-27.