LLaDA2.2 Released: Enabling Flexible Text Editing for Diffusion Language Models
jiqizhixin · x · 2026-08-09
Inclusion AI has released the LLaDA2.2 model, designed to allow AI agents to flexibly correct their own mistakes instead of rewriting everything from scratch.
The core breakthrough lies in introducing flexible text editing capabilities to diffusion language models. By comparing drafts with the correct output, the model can keep, substitute, delete, or insert tokens. Additionally, LLaDA2.2 incorporates reinforcement learning to improve decision-making from feedback, supports a 128K token context, and routes work efficiently across expert models. It matches or exceeds traditional autoregressive models on long-horizon agentic benchmarks.
More from Research
- Simplex Diffusion Models author Justin Deschenaux to present his paper Monday — LucaAmb · 2026-10-16
- Tool auto-audits paper claims with pass/fail evals — 4 of 7 metrics reproduce, 1 fails — jwt0625 · 2026-10-16
- Hironobu Fujiyoshi publishes deck: from AI agents to Agentic Robotics — HirokatuKataoka · 2026-10-16
- Is math just a language? A debate on why LLMs are good at mathematics — LucaAmb · 2026-10-16
- Alibaba open-sources OpenGCUA, a full-stack generalist computer-use agent; +19.6 points on OSWorld-v2 across 8 models — Tongyi-MAI · 2026-10-16
- CoDA co-evolves simulation and robot policies via shared failure memory, lifting OOD success from 5.3% to 27.6% — PLAN-Lab · 2026-10-16