List your company

Hallucination Detection on SciQ (AUROC, F1-Score)

0.9328AUROCCausalGaze
0.5076480.6180240.72840.838776Apr 13, 2026

Evaluation results

MethodLinks
CausalGazeLLM=Qwen2-7B2026.040.932882.83
HaluGNNLLM=Qwen2-7B2026.040.921784.02
CausalGazeLLM=Mistral-7B2026.040.85578.6
CausalGazeLLM=Llama2-7B2026.040.83977.01
HaluGNNLLM=Mistral-7B2026.040.834372.73
HaluGNNLLM=Llama2-7B2026.040.833676.49
SAPLMALLM=Qwen2-7B2026.040.81582
SAPLMALLM=Mistral-7B2026.040.788477.9
ICR-ProbeLLM=Mistral-7B2026.040.779378.26
ICR-ProbeLLM=Llama2-7B2026.040.774875.61
SAPLMALLM=Llama2-7B2026.040.77377.67
ICR-ProbeLLM=Qwen2-7B2026.040.759572.15
Semantic EntropyLLM=Mistral-7B2026.040.694566.4
Semantic EntropyLLM=Llama2-7B2026.040.68267.97
SelfCheckGPTLLM=Llama2-7B2026.040.67968.3
Semantic EntropyLLM=Qwen2-7B2026.040.65966.04
EigenScoreLLM=Mistral-7B2026.040.647463.46
EigenScoreLLM=Qwen2-7B2026.040.63264.62
LN-EntropyLLM=Mistral-7B2026.040.603658.32
EigenScoreLLM=Llama2-7B2026.040.59660.08
SelfCheckGPTLLM=Qwen2-7B2026.040.58657.79
LLM-CheckLLM=Llama2-7B2026.040.584256.37
LN-EntropyLLM=Llama2-7B2026.040.57658.26
LLM-CheckLLM=Mistral-7B2026.040.574855.47
LLM-CheckLLM=Qwen2-7B2026.040.572657.73
P(True)LLM=Mistral-7B2026.040.567355.31
SelfCheckGPTLLM=Mistral-7B2026.040.559352.6
PerplexityLLM=Mistral-7B2026.040.54755.83
P(True)LLM=Llama2-7B2026.040.54656.45
P(True)LLM=Qwen2-7B2026.040.53854.49
PerplexityLLM=Qwen2-7B2026.040.53454.77
PerplexityLLM=Llama2-7B2026.040.52652.56
LN-EntropyLLM=Qwen2-7B2026.040.52452.56