Evidence-Ledger Adjudication for Claim-Evidence Traceability
Source
arxiv.org
Author
Gengyu Chen, Yongjie Yu, Weiling Wang
Date
Why it matters
Pairing each generated claim with an evidence packet and routing unsupported ones back nearly doubles relation accuracy over non-AI agentAn AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.Full definition → baselines (0.676 vs 0.383) on blind labels from AVeriTeC, CLIMATE-FEVER and SciFact.
Terms in this piece · Glossary
AI agent — An AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.
articleECLoop: Evidence-Conditioned Execution Layer for Coding AgentsYisen Xu, Chenglin Li, Zehao Wang, Jinqiu Yang, Tse-Hsun Chen
articleEA-Graph: Artifact-Anchored Verification Memory for Coding Agents under Upstream DriftHwai-Jung Hsu, Cheng-Jan Chi, Hanna Everett
articleGetting the Source Right, Not Just the Fact: Source-Aware Verification for MCP Agentshuggingface.co