Loading…
Newsroom
Staleness-Adaptive Trust Region cuts asynchronous RL performance loss to 3% at 8× policy lagReleases · 7d agoDistilled RL transfers knowledge across model families without unconditional imitationResearch · 7d agoQwen-Music generates full vocal songs from text and lyricsReleases · 7d agoLongStraw trains RL models at 2.1M tokens on eight H20 GPUsReleases · 7d agoShortOPD cuts pruned LLM recovery time by 75% while raising generation quality 9×Releases · 7/15/2026Claude Design launches as Anthropic Labs visual collaboration toolPlatform · 7/14/2026