We have 2 of 30 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.
This paper identifies both the easiest and hardest to poison examples based on only clean model information, enabling both rigorous worst-case evaluation and proactive vulnerability-aware defense.
It is found that signal effectiveness is task-dependent: confidence is strongest on MCQ and simpler math, while likelihood/KL signals give the most frequent gains on harder math and code; no signal is universally best across model updates either, and some cross-version signals stay informative even when confidence fail...
We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.
