Visual-to-Symbolic Analytical Solution Inference on ViSA-Bench LLM-only setting
20Overall AccuracyGPT-5.2
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| GPT-5.2Input Mode=LLM-only2026.04 | 20 | 27.3 | 32.3 | 29.5 | |
| Claude-Haiku-4-5Input Mode=LLM-only2026.04 | 15.9 | 22.1 | 26.1 | 22.1 | |
| Grok-4-1-FastInput Mode=LLM-only2026.04 | 11.6 | 21.7 | 19.5 | 17.6 |