Using hallucinated benchmark answers as rejected DPO pairs, ordered by an external fact-checker's grounding score, improves hallucination detection in 1B-3B Llama models.
Title resolution pending
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
-
Teaching with Lies: Curriculum DPO on Synthetic Negatives for Hallucination Detection
Using hallucinated benchmark answers as rejected DPO pairs, ordered by an external fact-checker's grounding score, improves hallucination detection in 1B-3B Llama models.