Skip to main content
Search

Muhammad Zeeshan Karamat

1 related articles

Subscribe to updates
  1. arXiv

    Modifying only 0.19% of down_proj weights in LLaMA-2-7B-Chat raises attack success to 53% Basic ASR and 56% GCG ASR while tinyBenchmarks accuracy stays at 51.6%

    This work asks whether safety-sensitive behavior in LLaMA-2-7B-Chat is concentrated in a sparse subset of parameters, using low-rank safety-associated subspace analysis and parameter-level safety-utility importance filtering; both approaches show highly non-uniform safety sensitivity, with the MLP down_proj consistently the most prominent safety-sensitive component and o_proj a smaller contributor, and parameter-level localization modifying only 0.19% of model weights in down_proj yields 53% Basic ASR and 56% GCG ASR while tinyBenchmarks accuracy remains 51.6% against a 52.2% unmodified baseline.

You’re all caught up