List your company

Single-turn Dialogue on Anthropic HH

69.18Win RateIPO*
47.797653.348858.964.4512Feb 9, 2026Feb 22, 2026Mar 7, 2026Mar 21, 2026Apr 3, 2026Apr 16, 2026Apr 30, 2026

Evaluation results

MethodLinks
IPO*Temperature=0.72026.0269.18
TUR-DPOEvaluation Protocol=LLM judge2026.0467.9
PPOEvaluation Protocol=Human eval2026.0467.9
TUR-DPOEvaluation Protocol=Human eval2026.0467.2
IPOEvaluation Protocol=LLM judge2026.0466.1
DPOEvaluation Protocol=LLM judge2026.0465.5
IPO*Temperature=0.12026.0264.83
DPOEvaluation Protocol=Human eval2026.0464.7
DIPO+Temperature=0.72026.0261.34
DIPO+Temperature=0.12026.0260.53
DIPOTemperature=0.12026.0260.3
Sampled IPOTemperature=0.12026.0260.13
Sampled IPOTemperature=0.72026.0259.99
DIPOTemperature=0.72026.0258.75
IPOTemperature=0.72026.0257.38
IPOTemperature=0.12026.0257.19
SFTTemperature=0.72026.0252.77
SFTTemperature=0.12026.0248.62