AggTruth uses four ways of aggregating attention scores over the retrieved passage to train a logistic regression detector, achieving stable cross-task AUROC across four LLMs.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
AggTruth: Contextual Hallucination Detection using Aggregated Attention Scores in LLMs
AggTruth uses four ways of aggregating attention scores over the retrieved passage to train a logistic regression detector, achieving stable cross-task AUROC across four LLMs.