Hallucination Self-Play co-evolves a generator and detector from one base LLM via RLAIF and RLVR, lifting a 7B model to match advanced LLMs on RAGTruth faithfulness detection.
Verifying the verifiers: Unveiling pitfalls and potentials in fact verifiers.arXiv preprint arXiv:2506.13342
4 Pith papers cite this work. Polarity classification is still indexing.
4
Pith papers citing it
fields
cs.CL 4years
2026 4representative citing papers
A new framework parses and evaluates citations in LLM deep research reports across link validity, relevance, and factuality, finding 94%+ link success but only 39-77% factual accuracy.
Introduces claim-conditioned re-scoring (SIFT) and warranted supports proportion (WSP) metric, reporting accuracy recovery up to 27.6 points and WSP calibration at AUC 0.92 on FEVER, SciFact and other benchmarks.
SEVA trains a verification agent with a decomposed process reward to produce structured fact attributions, enabling a self-evolution loop that matches GPT-4o-mini F1 on ClearFacts while generating richer output.
citing papers explorer
-
Hallucination Self-Play: Bootstrapping Reinforced Detector via Evolved Generator
Hallucination Self-Play co-evolves a generator and detector from one base LLM via RLAIF and RLVR, lifting a 7B model to match advanced LLMs on RAGTruth faithfulness detection.
-
Cited but Not Verified: Parsing and Evaluating Source Attribution in LLM Deep Research Agents
A new framework parses and evaluates citations in LLM deep research reports across link validity, relevance, and factuality, finding 94%+ link success but only 39-77% factual accuracy.
-
The Warrant Gap: Claim-Conditioned Re-scoring for Fact-Checking
Introduces claim-conditioned re-scoring (SIFT) and warranted supports proportion (WSP) metric, reporting accuracy recovery up to 27.6 points and WSP calibration at AUC 0.92 on FEVER, SciFact and other benchmarks.
-
SEVA: Self-Evolving Verification Agent with Process Reward for Fact Attribution
SEVA trains a verification agent with a decomposed process reward to produce structured fact attributions, enabling a self-evolution loop that matches GPT-4o-mini F1 on ClearFacts while generating richer output.