AIInfoQabout 2 hours ago
Meta Researchers Launch Self-Managing Context Models
3 min read
Researchers from Meta, MIT, and the University of Washington introduce Context Language Models (CLMs) that allow language models to manage their own context. Zero-shot CLMs achieved 11.4% higher accuracy with 21.5% fewer FLOPs on BrowseComp-Plus. Reinforcement learning improved Qwen3.5-9B performance from 28.8% to 42.5% while using 12% fewer FLOPs.
Level
Hype check
Tap to vote and see what everyone thinks.
In this storyMeta