5 papers
You Had One Job: Per-Task Quantization Using LLMs' Hidden Representations
Amit LeVi, Raz Lapid, Rom Himelstein +3
Many LLM applications require only narrow capabilities, yet standard post-training quantization (PTQ) methods allocate precision without considering the target task. This can waste…
Jailbreak Attack Initializations as Extractors of Compliance Directions
Amit Levi, Rom Himelstein, Yaniv Nemcovsky +2
Safety-aligned LLMs respond to prompts with either compliance or refusal, each corresponding to distinct directions in the model's activation space. Recent works show that initiali…
Silenced Biases: The Dark Side LLMs Learned to Refuse
Rom Himelstein, Amit LeVi, Brit Youngmann +2
Safety-aligned large language models (LLMs) are becoming increasingly widespread, especially in sensitive applications where fairness is essential and biased outputs can cause sign…
Representing LLMs in Prompt Semantic Task Space
Idan Kashani, Avi Mendelson, Yaniv Nemcovsky
Large language models (LLMs) achieve impressive results over various tasks, and ever-expanding public repositories contain an abundance of pre-trained models. Therefore, identifyin…
Sparse patches adversarial attacks via extrapolating point-wise information
Yaniv Nemcovsky, Avi Mendelson, Chaim Baskin
Sparse and patch adversarial attacks were previously shown to be applicable in realistic settings and are considered a security risk to autonomous systems. Sparse adversarial pertu…