Models RSS

Agents 🇨🇳

World's First Agentic Diffusion Model: Self-Corrects While Acting, 128K Context Matches Autoregressive Models

Ant Group's inclusionAI team released LLaDA 2.2, the world's first large-scale agentic diffusion model with 128K native context. It integrates Levenshtein editing, reinforcement learning from environment feedback, and BlockRouting architecture, achieving performance close to top autoregressive models on agent benchmarks while offering higher throughput.

QbitAI 量子位28.07 · 09:03
Fresh news