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

Software Engineering on SWE-rebench 9B

38.16Avg@8 BC%

EvoTrainer

Updated 1mo ago

Evaluation Results

Method	Links
EvoTrainer 2026.06		38.16
RAGEN v2 SNR Filtering 2026.06		35.74
Human-engineered RL 2026.06		33.77
AutoResearch 2026.06		33.33
RAGEN v1 Filtering 2026.06		32.89
GRPO + Clip-Higher 2026.06		32.79
Group-Variance Filtering 2026.06		32.31
GSPO + Clip-Higher 2026.06		31.97
Base model (no RL) 2026.06		30.19