Preference Optimization Algorithm KTO Upgraded to Stable TRL API
ethayarajh · x · 2026-07-06
HuggingFace's training library TRL has officially promoted the preference learning method KTO (Kahneman-Tversky Optimization) to a stable API. KTOTrainer and KTOConfig have graduated from experimental interfaces, allowing researchers to conduct preference optimization training more reliably.
More from Research
- Could 10k agents discover learning methods beyond backprop, or just tweak existing ones? — SeunghyunSEO7 · 2026-09-11
- TRACES grades the process, not the answer: six-dimension eval for open-ended AI science — Faheem_uh · 2026-09-11
- Cognition's SWE-2 uses a KKT duality argument in RL to shift the effort Pareto curve — YouJiacheng · 2026-09-11
- VidMap uses RoMa coarse matching on all frames, fine-scale only for keyframes — ducha_aiki · 2026-09-11
- Bug Hunt Bench author: leaderboard noise is about 2-3 points — PawelHuryn · 2026-09-11
- PNAS paper shows a tiny billiard-ball system is a universal computer — undecidability lives in two dimensions — eigensteve · 2026-09-11