Papers2

#continual pretraining

Stable-DiffCoder: Pushing the Frontier of Code Diffusion Large Language Model

Chenghao Fan, Wen Heng et al.Jan 22arXiv

Stable-DiffCoder is a code-focused diffusion language model that learns to write and edit programs by filling in masked pieces, not just predicting the next token.

#diffusion language model#block diffusion#code generation

Not triaged yet

Let It Flow: Agentic Crafting on Rock and Roll, Building the ROME Model within an Open Agentic Learning Ecosystem

Intermediate

Weixun Wang, XiaoXiao Xu et al.Dec 31arXiv

This paper builds an open, end-to-end ecosystem (ALE) that lets AI agents plan, act, and fix their own mistakes across many steps in real computer environments.

#agentic LLMs#reinforcement learning#IPA

Not triaged yet