Investigating the Interplay between Contextual and Parametric Chain-of-Thought Faithfulness under Optimization
探究优化下上下文与参数化思维链忠实性之间的相互作用
机构 * University of Copenhagen(哥本哈根大学) ; Technische Universität Berlin(柏林技术大学) ; German Research Center for Artificial Intelligence (DFKI)(德国人工智能研究中心(DFKI)) ; BIFOLD – Berlin Institute for the Foundations of Learning and Data(BIFOLD – 柏林学习与数据基础研究院)
专题命中 推理评测 :chain-of-thought(title,abstract);CoT(abstract,abstract_cn);分类 cs.CL、cs.AI、cs.LG
AI总结 通过提出统一偏好对齐接口FaithMate,研究上下文与参数化两种思维链忠实性范式在优化下的相互作用,发现两者正相关但不对称,且上下文忠实性指标间存在权衡。
Comments The first two authors contributed equally and share first-authorship