List your company

Science on GPQA Diamond (Pass@1, Avg Token Length)

71.5Pass@1 ScoreDeepSeek-R1
61.72464.26266.869.338Mar 6, 2025

Evaluation results

MethodLinks
DeepSeek-R12025.0371.55,300
DeepSeek-R1-Distill-Llama-70BParameters=70B, Backbone=Llama2025.0365.2—
TinyR1-32B-PreviewParameters=32B2025.03658,600
DeepSeek-R1-Distill-Qwen-32BParameters=32B, Backbone=Qwen2025.0362.15,300