Skip to yearly menu bar Skip to main content


Oral
in
Workshop: Algorithmic Fairness through the lens of Metrics and Evaluation

Contributed talk: Evaluating and Mitigating Discrimination in Language Model Decisions

Alex Tamkin

[ ]
Sat 14 Dec 10:25 a.m. PST — 10:30 a.m. PST
 
presentation: Algorithmic Fairness through the lens of Metrics and Evaluation
Sat 14 Dec 9 a.m. PST — 5:30 p.m. PST

Abstract:

As language models (LMs) advance, interest is growing in applying them to high-stakes societal decisions, such as determining financing or housing eligibility. However, their potential for discrimination in such contexts raises ethical concerns, motivating the need for better methods to evaluate these risks. We present a method for proactively evaluating the potential discriminatory impact of LMs in a wide range of use cases, including hypothetical use cases where they have not yet been deployed. Specifically, we use an LM to generate a wide array of potential prompts that decision-makers may input into an LM, spanning 70 diverse decision scenarios across society, and systematically vary the demographic information in each prompt. Applying this methodology reveals patterns of both positive and negative discrimination across six different Claude models, in select settings when no interventions are applied. While we do not endorse or permit the use of language models to make automated decisions for the high-risk use cases we study, we demonstrate techniques to significantly decrease both positive and negative discrimination through careful prompt engineering, providing pathways toward safer deployment in use cases where they may be appropriate. Our work enables developers and policymakers to anticipate, measure, and address discrimination as language model capabilities and applications continue to expand. We release our dataset and prompts at [redacted for submission.

Chat is not available.