List your company

Truthfulness on TruthfulQA (avg.@8)

68.69Truthfulness Avg.@8PSFT
54.59858.256561.91565.5735Aug 25, 2025

Evaluation results

MethodLinks
PSFTBackbone=Llama3.1-8B-Instruct2025.0868.69
PSFTwarm-upBackbone=Llama3.1-8B-Instruct2025.0868.55
PSFTBackbone=Qwen2.5-7B-Instruct2025.0867.16
SFTBackbone=Llama3.1-8B-Instruct2025.0867.08
SFT-KLBackbone=Llama3.1-8B-Instruct2025.0867.03
PSFTwarm-upBackbone=Qwen2.5-7B-Instruct2025.0866.37
BaseBackbone=Qwen2.5-7B-Instruct2025.0866.1
SFTBackbone=Qwen2.5-7B-Instruct2025.0863.14
SFT-KLBackbone=Qwen2.5-7B-Instruct2025.0861.31
BaseBackbone=Llama3.1-8B-Instruct2025.0855.14