Mathematical Reasoning via Verifiable Subquestions on IMProofBench
71.8Subquestion ScoreGemini-3-Pro
Evaluation Results
| Method | Links | |
|---|---|---|
| Gemini-3-Pro2026.02 | 71.8 | |
| GPT-52026.02 | 54.5 | |
| Claude-Opus-4.12026.02 | 38.7 |
| Method | Links | |
|---|---|---|
| Gemini-3-Pro2026.02 | 71.8 | |
| GPT-52026.02 | 54.5 | |
| Claude-Opus-4.12026.02 | 38.7 |