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

Scientific Reasoning on GPQA (pass@1)

70.2pass@1

EvoEnv

Updated 19d ago

Evaluation Results

Method	Links
EvoEnv 2026.05		70.2
RLVE 2026.05		67.7
R-Zero 2026.05		67.2
RLVE 2026.05		66.2
EvoEnv 2026.05		65.7
Untrained 2026.05		65.2
Untrained 2026.05		65.2
DAPO 2026.05		64.6
DAPO 2026.05		61.6
R-Zero 2026.05		58.1
RLVE 2026.05		57.1
TTVS 2026.04		56.1
DAPO 2026.05		56.1
EvoEnv 2026.05		55
R-Zero 2026.05		54
TTRL 2026.04		53
Untrained 2026.05		52.5
TTVS 2026.04		51.5
TTVS 2026.04		48.9
TTRL 2026.04		48.5
Qwen3-8B 2026.04		48
TTRL 2026.04		43
TTVS 2026.04		37.3
TTVS 2026.04		36.4
Qwen2.5-Instruct-7B 2026.04		36.4
TTRL 2026.04		34.1
TTRL 2026.04		33.3
LLaMA3.1-8B 2026.04		30.8
Qwen2.5-Instruct-3B 2026.04		29.3
OpenReasoner-Zero-7B 2026.04		28.4
TTVS 2026.04		28.3
SimpleRL-Zero-7B 2026.04		27.6
TTVS 2026.04		26.3
Qwen3-4B 2026.04		26.3
DeepSeek-R1-Distill-7B 2026.04		25.7
TTRL 2026.04		25.3
TTRL 2026.04		22.2
TTVS 2026.04		17.7
Qwen2.5-Instruct-1.5B 2026.04		16.7
TTRL 2026.04		16.1
DeepSeek-R1-Distill-1.5B 2026.04		14.6
Qwen3-1.7B-Instruct 2026.04		13.1
LLaMA-3.2-3B-Instruct 2026.04		12.5