探究大模型内部如何表征代码正确性，检验其推理鲁棒性的前沿研究。
arXiv:2608.08266v1 Announce Type: cross Abstract: Code generated by modern language models often reads naturally. Yet, it also often fails to implemen…
揭秘大模型安全软肋：扰动型越狱提示的几何结构，小模型内部表征泄露攻击机制。
arXiv:2607.20581v1 Announce Type: cross Abstract: Perturbation techniques that turn unsuccessful jailbreak prompts into successful ones are continuous…
大模型幻觉信号能否跨语言、跨领域通用？这项研究直接检验内部表征的泛化能力，为幻觉检测提供新视角。
arXiv:2607.04029v1 Announce Type: new Abstract: Recent hallucination detection techniques in large language models (LLMs) focus on directly extracting…
大语言模型秘密编码剩余输出长度，揭示其生成结构的可预测线性规律。
arXiv:2607.05316v1 Announce Type: cross Abstract: Large language models generate one token at a time, yet their responses show remarkably consistent l…
挑战Anthropic关于LLM具有“功能性情感”的断言，从生物系统视角剖析情感本质，引发对AI情感能力的深度思辨。
arXiv:2606.14742v1 Announce Type: cross Abstract: Do LLMs have emotions? A recent paper from Anthropic reports finding internal representations of emo…
揭示LLM临床分诊失败根源：内部表征问题而非临床知识不足，输出格式影响评估结果
arXiv:2605.29889v1 Announce Type: cross Abstract: Patient-voiced clinical-triage benchmarks report high under-triage rates for consumer LLMs for const…
突破线性干预局限，提出大模型非线性特征干预新方法，为理解LLM内部表征开辟新路径
arXiv:2605.14749v1 Announce Type: cross Abstract: Intervention is one of the most representative and widely used methods for understanding the interna…