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

Physical Reasoning on PHYRE DeepPHY (test)

14.92Att. 1 Score

Qwen-7B^P&I GRPO

Updated 1mo ago

Evaluation Results

Method	Links
Qwen-7B^P&I GRPO 2026.03		14.92	45.56
Qwen-7B^P&I GRPO 2026.03		14.63	43.17
Qwen-3B^P&I GRPO 2026.03		14.11	39.49
Qwen-3B^P&I GRPO 2026.03		14	34.5
Qwen-7B^P GRPO 2026.03		13.18	43.66
Qwen-7B^P GRPO 2026.03		13	40
Qwen-3B^P&I SFT 2026.03		10.7	33.56
Qwen-3B^P&I SFT 2026.03		10.42	29
Qwen-3B^P GRPO 2026.03		9.85	32
Qwen-3B^P GRPO 2026.03		9.5	29.03
Qwen-7B^P&I SFT 2026.03		7.95	44.85
Qwen-7B^P&I SFT 2026.03		7.5	42.67
Qwen-7B^P SFT 2026.03		3.67	36.13
Qwen-7B^P SFT 2026.03		3.4	38.67
GPT-o3 2026.03		3.03	30.77
Qwen-72B 2026.03		2.43	14.92
Gemini-2.5-Pro 2026.03		2.17	22.07
Gemini-2.5-Pro 2026.03		2.17	12.33
Qwen-72B 2026.03		1.78	10.39
Claude 4.0 Opus 2026.03		1.73	10.63
Qwen-3B^P SFT 2026.03		1.52	28.69
MOCK 2026.03		1.5	10.8
Qwen-3B^P SFT 2026.03		1.33	25.58
Claude 4.0 Opus 2026.03		1.11	7.44
Qwen-7B^I SFT 2026.03		0.82	13.45
Qwen-3B^I SFT 2026.03		0.74	11.88
Qwen-7B 2026.03		0.67	10.1
Qwen-7B^I SFT 2026.03		0.67	9.33
Qwen-7B^I GRPO 2026.03		0.67	8.33
Qwen-7B^I GRPO 2026.03		0.67	13.04
Qwen-3B^I GRPO 2026.03		0.66	13.48
Qwen-3B^I SFT 2026.03		0.33	9.83
Qwen-3B 2026.03		0.17	5.85
GPT-o3 2026.03		0.17	25.6
Qwen-3B^I GRPO 2026.03		0.17	9.67
Qwen-32B 2026.03		0.03	8.7