Tag: Chain of Thought
All the papers with the tag "Chain of Thought".
Crosslingual Reasoning through Test-Time Scaling
grok-3-latestScore: 0.89Published: at 16:50本文通过测试时计算扩展,揭示了英语中心推理模型在多语言数学推理中的潜力,并分析了语言混合模式和跨领域泛化的局限,为多语言推理研究提供了重要基准。
Optimizing Chain-of-Thought Reasoners via Gradient Variance Minimization in Rejection Sampling and RL
grok-3-latestScore: 0.58Published: at 06:26本文提出GVM-RAFT方法,通过动态采样分配策略最小化梯度方差,显著提升大型语言模型在链式思维推理任务中的训练效率和性能。
Between Underthinking and Overthinking: An Empirical Study of Reasoning Length and correctness in LLMs
grok-3-latestScore: 0.86Published: at 18:48本文通过实证分析揭示了大型语言模型推理长度与正确性之间的非线性关系,并通过长度偏好优化实验提出了一种在无监督条件下减少生成长度的有效方法,为自适应推理研究提供了新视角。