Se-DPO assigns evolving per-token credits from implicit reward magnitude and reference entropy during DPO training, improving instruction-following win rates but with an internally inconsistent derivation.
Kl penalty control via perturbation for direct preference optimization.arXiv preprint arXiv:2502.13177,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CL 1years
2026 1verdicts
REJECT 1roles
background 1polarities
background 1representative citing papers
citing papers explorer
-
Se-DPO: Self-Evolving Token Credit for Direct Preference Optimization
Se-DPO assigns evolving per-token credits from implicit reward magnitude and reference entropy during DPO training, improving instruction-following win rates but with an internally inconsistent derivation.