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

Physical Reasoning on I-PHYRE DeepPHY (test)

37.6Attention Score 1

Qwen-3B^I GRPO

Updated 1mo ago

Evaluation Results

Method	Links
Qwen-3B^I GRPO 2026.03		37.6	51
Claude 4.0 Opus 2026.03		36.67	56.67
Qwen-3B^I GRPO 2026.03		36.67	48.89
Qwen-3B^P&I GRPO 2026.03		36.67	60
Qwen-3B^P&I GRPO 2026.03		35.66	61.99
Qwen-7B 2026.03		32.42	32.42
Claude 4.0 Opus 2026.03		30	50
GPT-o3 2026.03		30	86.67
Qwen-7B^I GRPO 2026.03		27.82	89.31
Qwen-7B^I GRPO 2026.03		26.67	86.67
GPT-o3 2026.03		25	76.67
MOCK 2026.03		23.33	53.33
Qwen-3B^I SFT 2026.03		23.33	45.56
Qwen-3B^I SFT 2026.03		23.21	47.34
Qwen-7B^P&I SFT 2026.03		21.41	65.38
Qwen-3B^P&I SFT 2026.03		21.29	52.91
Gemini-2.5-Pro 2026.03		20	63.33
Qwen-3B^P SFT 2026.03		20	30
Qwen-3B^P&I SFT 2026.03		20	50
Qwen-7B^P&I SFT 2026.03		20	63.33
Qwen-3B^P SFT 2026.03		19.87	32.45
Gemini-2.5-Pro 2026.03		16.67	53.33
Qwen-7B^P GRPO 2026.03		16.67	28.89
Qwen-7B^P GRPO 2026.03		15.68	30
Qwen-7B^P SFT 2026.03		13.57	22.15
Qwen-3B 2026.03		13.33	16.67
Qwen-72B 2026.03		13.33	43.33
Qwen-3B^P GRPO 2026.03		13.33	26.67
Qwen-7B^P SFT 2026.03		13.33	19.33
Qwen-7B^P&I GRPO 2026.03		13.33	90
Qwen-7B^P&I GRPO 2026.03		13.12	93.33
Qwen-3B^P GRPO 2026.03		12.78	28
Qwen-7B^I SFT 2026.03		12.22	50
Qwen-7B^I SFT 2026.03		11.34	51.5
Qwen-72B 2026.03		10	36.67
Qwen-32B 2026.03		0.17	5.6