List your company

Graduate-Level Q&A on GPQA diamond en

33.87AccuracyTLPO
27.6329.2530.8732.49Apr 29, 2026

Evaluation results

MethodLinks
TLPOFine-tuning=After fine-tuning2026.0433.87
BaselineFine-tuning=Prior to fine-tuning2026.0432.54
DPOFine-tuning=After fine-tuning2026.0431.53
ORPOFine-tuning=After fine-tuning2026.0431.06
SFTFine-tuning=After fine-tuning2026.0427.87