Share your thoughts, 1 month free Claude Pro on usSee more

Multiple-Choice Question Answering on PhyX 1k (test)

79.7MCQ Exact Accuracy

Claude Sonnet 4.5

Updated 19d ago

Evaluation Results

Method	Links
Claude Sonnet 4.5 2026.05		79.7
Physics-R1 (dense) 2026.05		78.3
Physics-R1 (binary, seed 42) 2026.05		78
Physics-R1 (binary, 3-seed mean ±σ) 2026.05		77.8
Gemini 2.5 Pro 2026.05		75.1
Qwen3-VL-32B-Thinking 2026.05		73.8
Qwen3-VL-8B-Thinking (base) 2026.05		73.7
GPT-4o 2026.05		70.4
InternVL3-8B 2026.05		46.8