Preference Learning with Response Time: Robust Losses and Guarantees

TL;DR

This paper introduces a novel approach to preference learning that incorporates response time data, improving reward model accuracy and efficiency through new loss functions and theoretical guarantees, especially for complex reward functions.

Contribution

It proposes response time-augmented preference learning methods with Neyman-orthogonal losses, providing theoretical guarantees and improved sample efficiency over traditional approaches.

Findings

Response time data enhances reward model learning.

The proposed methods achieve oracle convergence rates.

Significant reduction in error scaling from exponential to polynomial.

Abstract

This paper investigates the integration of response time data into human preference learning frameworks for more effective reward model elicitation. While binary preference data has become fundamental in fine-tuning foundation models, generative AI systems, and other large-scale models, the valuable temporal information inherent in user decision-making remains largely unexploited. We propose novel methodologies to incorporate response time information alongside binary choice data, leveraging the Evidence Accumulation Drift Diffusion (EZ) model, under which response time is informative of the preference strength. We develop Neyman-orthogonal loss functions that achieve oracle convergence rates for reward model learning, matching the theoretical optimal rates that would be attained if the expected response times for each query were known a priori. Our theoretical analysis demonstrates that for linear reward functions, conventional preference learning suffers from error rates that scale exponentially with reward magnitude. In contrast, our response time-augmented approach reduces this to polynomial scaling, representing a significant improvement in sample efficiency. We extend these guarantees to non-parametric reward function spaces, establishing convergence properties for more complex, realistic reward models. Our extensive experiments validate our theoretical findings in the context of preference learning over images.