Kahneman-Tversky Optimization (KTO)
An alignment algorithm that optimizes models using binary utility signals (good/bad) rather than pairwise preferences.
KTO aligns language models based on Prospect Theory, evaluating individual generations as positive or negative utilities. This eliminates the need for paired comparison data, simplifying dataset curation.
Historical figures and technical concepts for informational purposes only. Not technical, professional, legal, or financial advice. Sources: Official Documentation.