2 papers
cs.SE2026
Diff Diffusion: Predicting Edit Paths with Diffusion Models for Efficient Code Editing
Xiaolong Xu, Hao Jiang, Bo Jiang +5
Diffusion large language models (dLLMs) are appealing for code generation because they can update many tokens in parallel, yet many are optimized for sequence synthesis rather than…
cs.LG2026
LFPO: Likelihood-Free Policy Optimization for Masked Diffusion Models
Chenxing Wei, Jiazhen Kang, Hong Wang +8
Reinforcement Learning with Verifiable Rewards (RLVR) has achieved remarkable success in improving autoregressive models, especially in domains requiring correctness like mathemati…