cs.LGFeb 2, 2024

KTO: Model Alignment as Prospect Theoretic Optimization

Authors: Kawin EthayarajhWinnie XuNiklas MuennighoffDan JurafskyDouwe Kiela

Abstract

Kahneman & Tversky's prospect theory\textit{prospect theory} tells us that humans perceive random variables in a biased but well-defined manner (1992); for example, humans are famously loss-averse. We show that objectives for aligning LLMs with human feedback implicitly incorporate many of these biases -- the success of these objectives (e.g., DPO) over cross-entropy minimization can partly be ascribed to them belonging to a family of loss functions that we call human-aware losses\textit{human-aware losses} (HALOs). However, the utility functions these methods attribute to humans still differ from those in the prospect theory literature. Using a Kahneman-Tversky model of human utility, we propose a HALO that directly maximizes the utility of generations instead of maximizing the log-likelihood of preferences, as current methods do. We call this approach KTO, and it matches or exceeds the performance of preference-based methods at scales from 1B to 30B, despite only learning from a binary signal of whether an output is desirable. More broadly, our work suggests that there is no one HALO that is universally superior; the best loss depends on the inductive biases most appropriate for a given setting, an oft-overlooked consideration.

Explore similar work

CardsList
  1. HAL: Inducing Human-likeness in LLMs with Alignment

    Jan 6, 2026Masum Hasan, Junjie Zhao, Ehsan HoqueValue Alignment