Reproduce “Hallucination Across the Reasoning Lifecycle: Interface Visibility, Causal Evidence, and Release Control in Large Reasoning Models” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Zhe Yu, Mohan Li, Lei Yu, Ka-Ho Chow, Chengwei Qin, Xingyu Wu, Wenpeng Xing, Shuguang Xiong and 1 more
The authors synthesize UIPCA, a diagnostic framework for tracing unsupported premises, invalid inferences, dependent reuse of confirmed errors, visible answer–trace contradictions, and policy violations. The framework links diagnosis to repair and release actions, while treating judgments as dependent on available evidence and interface visibility. Among 34 corpus sources and 24 external sources reviewed, none establishes the specified matched reasoning–reliability trade-off.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.