#hidden states

4 stories · 0 in 7 days · 0 in 30 days

LLM 1 transformer 1 τ-Bench 1 language models 1 logical reasoning 1

Stories, products and papers tagged "hidden states", newest first.

Sep 3

PapersSelected73°09:07

Logical Validity Representations in Language Models

Models can represent logical validity internally without reliably expressing it in outputs, affecting how we evaluate AI reasoning capabilities.

#transformer #logical reasoning #language models #hidden states

arXiv cs.LG@Smitha Muthya Sudheendra, Jaideep SrivastavaSource

Sep 1

PapersSelected73°10:17

Hidden-State Trajectories for Adaptive Reasoning

This method improves LLM reasoning efficiency across tasks, reducing costs while increasing success rates, making it valuable for developers working with constrained resources.

#LLM #hidden states #τ-Bench

arXiv cs.AI@Jie Liang et al. (4)Source

Jul 28

Papers12:10

Chinese contentD-Score：大语言模型中基于谱分析的隐藏状态幻觉检测信号

这篇论文给出一个简单检测幻觉的方法，只靠模型内部的隐藏状态算一个分数，不用外部工具，在FAVA和RAG上验证有效。

#D-Score #FAVA-Annotation #RAGTruth #hallucination detection

arXiv cs.AI@Bianca Raimondi et al. (4)Source

Jun 15

Papers11:11

Chinese contentQwen3代码正确性在隐藏状态中的可解码性

论文揭示Qwen3隐藏状态如何预测代码正确性

#Qwen3 #code correctness #hidden states #LiveCodeBench

arXiv cs.LG@Carlo Di CiccoSource