Zhuo-Xin Liu

RiskChainBench introduces RiskChainBench, pairing 3,600 synthetic token-text restoration inputs from 600 source sessions with 600 corresponding human-labeled local web environments that produces a frozen, evidence-cited risk report without message-side semantics or domain-reputation cues.

Models struggle with handwritten text, showing a 63-91% reliance on language priors for errors, while humans exhibit a more balanced error profile.

Achieving a Relational Delivery Score of 66.6% reveals that even advanced models struggle with the complexities of merging interacting pull requests safely.