Author name: Wanqing Cui, Wei Huang, Keping Bi, Jiafeng Guo, Xueqi Cheng

arXiv:2502.13464v2 Announce Type: replace
Abstract: Commonsense plausibility estimation is critical for evaluating language models (LMs), yet existing generative approaches–reliant on likelihoods or verbalized judgments–struggle with fine-grained di…