A courtroom-style multi-agent debate with progressive retrieval reaches 81.7% accuracy on Check-COVID binary claims, 10 points above a simple MAD baseline, though a single-call RAG baseline outperforms it.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Courtroom-Style Multi-Agent Debate with Progressive RAG and Role-Switching for Controversial Claim Verification
A courtroom-style multi-agent debate with progressive retrieval reaches 81.7% accuracy on Check-COVID binary claims, 10 points above a simple MAD baseline, though a single-call RAG baseline outperforms it.