IDRF: Inverse-Distilled Reward Fine-tuning of Masked Discrete Diffusion Models

Paper recorded by Signals 4 on 2026-10-02 in cs.LG. Abstract reproduced from arXiv; link to the original below.

Published 2026-10-02 on arXiv · recorded by Signals 4 on 2026-10-05

Category: cs.LG · 机器学习 · first seen 2026-10-05

Abstract

Masked discrete diffusion models offer a promising alternative to autoregressive generation, but iterative sampling can be costly, and intractable sequence likelihoods complicate reward fine-tuning. We introduce IDRF, a framework for reward fine-tuning of few-step masked discrete diffusion generators. Starting from a standard reverse-KL-regularized objective, IDRF replaces the intractable sequence

Read on arXiv →