List your company

Expert Reasoning on GPQA

57.3ScoreDR-Rubric-8B (Gemini)
38.26843.20948.1553.091May 31, 2026

Evaluation results

MethodLinks
DR-Rubric-8B (Gemini)Training=SFT+RL, 1K2026.0557.3
WebExplorer-8BTraining=SFT+RL, 25K2026.0556
DR-Rubric-8B (BS-3)Training=SFT+RL, 3K2026.0555.8
DR-Rubric-8B (GPT-5)Training=SFT+RL, 1K2026.0554
Qwen2.5-7B2026.0552.3
Qwen3-8B-SFTTraining=SFT, 1K2026.0550.8
DR-Tulu-RL-8BTraining=SFT+RL, 25K2026.0544
DR-Tulu-SFT-8BTraining=SFT, 16K2026.0542
Qwen3-8B2026.0540
Search-R1-7BTraining=RL, 90K2026.0539