List your company

Hallucination Correction on TruthfulQA gated mode (test)

77.9IGR (%)DoLa
-3.11617.91738.9559.983May 7, 2026

Evaluation results

MethodLinks
DoLaModel=Llama-3.2-1B, Evaluation Mode=Gated mode2026.0577.94454264
ITIModel=Llama-3.2-1B, Evaluation Mode=Gated mode2026.0577.91926479
AdaSteerModel=Llama-3.2-1B, Evaluation Mode=Gated mode2026.0577.95223416
SADIModel=Llama-3.2-1B, Evaluation Mode=Gated mode2026.0577.95123417
ICDModel=Llama-3.2-1B, Evaluation Mode=Gated mode2026.0577.95127447
PC-LDCDModel=Llama-3.2-1B, Evaluation Mode=Gated mode2026.0577.950365114
DoLaModel=Mistral-7B, Evaluation Mode=Gated mode2026.0558.36971297
ITIModel=Mistral-7B, Evaluation Mode=Gated mode2026.0558.344314710
AdaSteerModel=Mistral-7B, Evaluation Mode=Gated mode2026.0558.369345013
SADIModel=Mistral-7B, Evaluation Mode=Gated mode2026.0558.369345013
ICDModel=Mistral-7B, Evaluation Mode=Gated mode2026.0558.364355115
PC-LDCDModel=Mistral-7B, Evaluation Mode=Gated mode2026.0558.372547229
DoLaModel=Qwen3-4B, Evaluation Mode=Gated mode2026.0545.47276256
ITIModel=Qwen3-4B, Evaluation Mode=Gated mode2026.0545.460355211
AdaSteerModel=Qwen3-4B, Evaluation Mode=Gated mode2026.0545.475335011
SADIModel=Qwen3-4B, Evaluation Mode=Gated mode2026.0545.476324910
ICDModel=Qwen3-4B, Evaluation Mode=Gated mode2026.0545.473365311
PC-LDCDModel=Qwen3-4B, Evaluation Mode=Gated mode2026.0545.478516823
DoLaModel=Llama-3.1-8B, Evaluation Mode=Gated mode2026.0541.46655293
ITIModel=Llama-3.1-8B, Evaluation Mode=Gated mode2026.0541.446325010
AdaSteerModel=Llama-3.1-8B, Evaluation Mode=Gated mode2026.0541.46729448
SADIModel=Llama-3.1-8B, Evaluation Mode=Gated mode2026.0541.46629448
ICDModel=Llama-3.1-8B, Evaluation Mode=Gated mode2026.0541.469375315
PC-LDCDModel=Llama-3.1-8B, Evaluation Mode=Gated mode2026.0541.469516525
VanillaModel=Llama-3.2-1B, Evaluation Mode=Gated mode2026.0505023416
VanillaModel=Qwen3-4B, Evaluation Mode=Gated mode2026.05050335011
VanillaModel=Mistral-7B, Evaluation Mode=Gated mode2026.05050345013
VanillaModel=Llama-3.1-8B, Evaluation Mode=Gated mode2026.0505029448