arXiv AI By Eunna Lee

Adaptive Capitulation: A Structural Failure Mode of LLM Responses in Vulnerability Contexts

Read the original on arXiv AI →

arXiv:2607. 19629v1 Announce Type: cross Abstract: Large language models operating in emotionally sensitive contexts face a structural trilemma: when users in vulnerable states request information that may reinforce maladaptive attribution, current response architectures resolve the tension through protective restriction, uninflected facilitation, or unintegrated co-presence of both imperatives -- each preserving one objective at the cost of the other.

Summary generated by The Flow from the publisher's feed. The full article lives at arXiv AI.

arXiv AI
Jul 28

Do LLMs Know Their Vulnerable Scenarios?

arXiv:2607. 23496v1 Announce Type: new Abstract: Safety-aligned large language models are trained to refuse harmful requests, yet embedding the same requests in particular scenarios can bypass their safeguards.

By Ziheng Peng, Huiqi Deng, Haoran Jing, Xuankun Rong, Jiahui Han, Xiting Wang, Na Zou, Xia Hu
Hugging Face Trending Papers
Jul 26

Do LLMs Know Their Vulnerable Scenarios?

Safety-aligned large language models are trained to refuse harmful requests, yet embedding the same requests in particular scenarios can bypass their safeguards. Existing red-teaming methods empirically identify effective scenarios through observed attack outcomes, but why particular scenarios weaken refusal remains mechanistically unclear.

arXiv AI
Jun 9

When Behavioral Safety Evaluation Fails: A Representation-Level Perspective

arXiv:2606. 08044v1 Announce Type: cross Abstract: Large Language Model (LLM) safety has often been evaluated at the behavior level, which provides limited evidence of internal robustness, as these evaluations target outputs rather than representation-level vulnerability under intervention.

By Enyi Jiang, Anders Gj{\o}lbye, Yibo Jacky Zhang, Sanmi Koyejo