We have 6 of 20 papers
We haven’t gathered this author’s papers yet. Follow them and we’ll fetch their work.
Not the right person? Other researchers publish under this name.
Safety interpretability advances the study of Large Language Model (LLM) alignment from behavioral constraints driven by data or algorithms towards a deeper understanding of internal mechanisms. However, existing works have focused primarily on safety-related representations, attention heads, or neurons after alignment...
Visual token compression reduces the inference cost of Large Vision-Language Models (LVLMs). However, aggregate robustness measures do not reveal whether a particular adversarial failure is induced by compression or inherited from the underlying model. We define a compression-specific failure (CSF) as an adversarial in...
This work proposed PLUME, a training-free method that constructs a global update representation, activates memory evidence to form a local parameter view, and adaptively integrates their predictions during decoding, which demonstrated its effectiveness in sequential evolution settings.
GRAFT, an IPP-oriented post-training framework combining constraint-aware SFT and preference optimization is proposed, an IPP-oriented post-training framework combining constraint-aware SFT and preference optimization that raises the average IIF score and preserves general generation ability.
Although text-to-image diffusion models exhibit remarkable generative power, concept erasure techniques are essential for preventing harmful content. Existing adversarial probes evaluate these methods by testing whether erased concepts can still be recovered. However, existing erasure and probe methods remain largely t...
This survey examines the protective paradigm that has grown around this intervention point, and finds that most protections are still validated mainly against static or weakly adaptive adversaries, while evidence beyond controlled benchmarks remains scarce.
We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.
