文章预览
LG - 机器学习 CV - 计算机视觉 CL - 计算与语言 1、[CL] Forest-of-Thought：Scaling Test-Time Compute for Enhancing LLM Reasoning 2、[LG] Test-Time Alignment via Hypothesis Reweighting 3、[CL] Large Concept Models：Language Modeling in a Sentence Representation Space 4、[CL] LatentQA：Teaching LLMs to Decode Activations Into Natural Language 5、[LG] Does Representation Matter? Exploring Intermediate Layers in Large Language Models 摘要：用思维森林扩展测试时计算以增强LLM推理、基于假设重加权的测试时对齐、在句子表示空间进行语言建模的大型概念模型、教大语言模型将激活解码为自然语言、大型语言模型中间层探索 1、[CL] Forest-of-Thought: Scaling Test-Time Compute for Enhancing LLM Reasoning Z Bi, K Han, C Liu, Y Tang... [Huawei Noah’s Ark Lab] 思维森林：扩展测试时计算以增强LLM推理 要点: 单纯增加树状思维（ToT）方法中叶子节点的数量，并不能无限提高
………………………………
