2 citations · 12 across the 17 of their papers we have counts for
5 papers · 1 filter
Comparison requires valid measurement: Rethinking attack success rate comparisons in AI red teaming
Alexandra Chouldechova, A. Feder Cooper, Solon Barocas +3
We argue that conclusions drawn about relative system safety or attack method efficacy via AI red teaming are often not supported by evidence provided by attack success rate (ASR)…
Validating LLM-as-a-Judge Systems under Rating Indeterminacy
Luke Guerdan, Solon Barocas, Kenneth Holstein +3
The LLM-as-a-judge paradigm, in which a judge LLM system replaces human raters in rating the outputs of other generative AI (GenAI) systems, plays a critical role in scaling and st…
Machine Unlearning Doesn't Do What You Think: Lessons for Generative AI Policy and Research
A. Feder Cooper, Christopher A. Choquette-Choo, Miranda Bogen +34
"Machine unlearning" is a popular proposed solution for mitigating the existence of content in an AI model that is problematic for legal or moral reasons, including privacy, copyri…
A Framework for Evaluating LLMs Under Task Indeterminacy
Luke Guerdan, Hanna Wallach, Solon Barocas +1
Large language model (LLM) evaluations often assume there is a single correct response -- a gold label -- for each item in the evaluation corpus. However, some tasks can be ambiguo…
SureMap: Simultaneous Mean Estimation for Single-Task and Multi-Task Disaggregated Evaluation
Mikhail Khodak, Lester Mackey, Alexandra Chouldechova +1
Disaggregated evaluation -- estimation of performance of a machine learning model on different subpopulations -- is a core task when assessing performance and group-fairness of AI…