Mosaic

Headlines · 2

1. 2d
   Mosaic: multi-session agentic reinforcement learning

   A paper submitted to arXiv on 15 Oct 2026, "Mosaic: Multi-Session Agentic Reinforcement Learning," introduces multi-session agentic RL, a more general setting in which a trajectory may span an arbitrary number of sessions to match inference-time execution such as context compaction and subagent forking. The authors note that the common practice of broadcasting trajectory-level advantages to all sessions neither distinguishes session contributions nor guides session allocation, so they propose Mosaic, a critic-free hierarchical method that folds session-usage signals into trajectory-level advantage estimation and shapes each session's advantage by its estimated marginal contribution. Experiments on BrowseComp with Qwen3.5-4B show Mosaic outperforming broadcasting in both context compaction and subagent forking scenarios while using fewer sessions on average.

   Hugging Face · Papers · 🔥 0
2. Oct 5
   Understanding AI introduces new staff writer Dan Kagan-Kans

   The AI newsletter Understanding AI published the first post by its newest writer, Dan Kagan-Kans, who spent a decade as managing editor of the magazine Mosaic before turning to freelance AI journalism. His recent work includes a New York Magazine profile of Hans Moravec and an analysis of the data center backlash. He is supported by a fellowship from the Tarbell Center for AI Journalism, which says fellows retain complete autonomy over their reporting.

   Understanding AI · 🔥 0

Experience and discussion from the community

Share my Mosaic experienceAsk about Mosaic