List your company

Hallucination Detection (Dialogue) on HaluEval DA

77.1F1 ScoreHalluClean
16.98832.59448.263.806Nov 12, 2025

Evaluation results

MethodLinks
HalluCleanBackbone=DeepSeek-V32025.1177.172.5
GPT-4o-miniStrategy=Direct Ask2025.1176.574
DeepSeek-V3Strategy=Direct Ask2025.1175.471.8
HalluCleanBackbone=GPT-3.5-turbo2025.1174.369.3
HalluCleanBackbone=Llama-3-70B2025.117468.8
ChatProtectBackbone=GPT-3.5-turbo2025.117269.3
DeepSeek-R1Strategy=Direct Ask2025.117169.8
Llama-3-70BStrategy=Direct Ask2025.1166.462.3
GPT-3.5-turboStrategy=Direct Ask2025.1162.866
Step-by-StepBackbone=GPT-3.5-turbo2025.1161.465.8
SelfCheckGPTBackbone=GPT-3.5-turbo2025.1119.927.8
Plan-and-SolveBackbone=GPT-3.5-turbo2025.1119.352