This work introduces TranSGrid, a testbed that brings deductive, inductive, and abductive reasoning together within a unified task and introduces two simplifications into TranSGrid, showing that existing tasks reduce either or both of the inductive and abductive demands of systematic generalization.
Cheng-Wen Qi, De-Heng Ye, Ya-Tao Bian· 0 citations
The results show that chemical CoT is neither a faithful explanation nor merely a post-hoc rationalization, but a hallucination-prone molecular scratchpad, which cautions against treating CoT as direct evidence of faithful reasoning and motivates process-level supervision beyond answer-only evaluation.
Jiatong Li, Yuxuan Ren, Weida Wang et al.· arXiv.org· 1 citation
A Molecular Perturbation framework that generates syntax-valid structural variants of training molecules under controlled Graph Edit Distance (GED) to probe the manifold regularity of molecular LLMs and suggests that it can partially expand the local trust region and offer a promising direction for stabilizing molecula...
Jiatong Li, Weida Wang, Changmeng Zheng et al.· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.