PIQA (PapersWithCode): leaderboard
PIQA, Physical Interaction: Question Answering (Bisk et al., 2020): choose the better of two solutions for reaching an everyday physical goal. This board is the archived Papers With Code table of accuracies as each paper reported them, under that paper’s own setup, not a common evaluation run; higher is better.
Metric: Accuracy. Source: huggingface.co. Saturation forecast: Rough model projection: around 2026. 66 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | OPT-175B | 81.07 | #1507 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=piqa-paperswithcode · How It Works · Data refreshed daily, snapshot 2026-11-05.