CLATTER, a structured decompose-attribute-aggregate reasoning prompt, improves NLI-based hallucination detection for reasoning models by about 3.8 accuracy points on average, with inconsistent results across models and datasets.
Prompt 1.3: Decomposition-Based Instructions Follow the steps below to guide your assessment:
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
CLATTER: Comprehensive Entailment Reasoning for Hallucination Detection
CLATTER, a structured decompose-attribute-aggregate reasoning prompt, improves NLI-based hallucination detection for reasoning models by about 3.8 accuracy points on average, with inconsistent results across models and datasets.