Meta Introduces Context Language Models (CLM), Open-Sourced

Meta, with UW and MIT, proposed Context Language Models (CLM), which treat context as a freely editable file, boosting agent task scores by 65% at equal compute while saving 21.5% compute and gaining 11.4% accuracy. Paper and code are open-sourced.

2026-09-30 ~ 2026-09-30 · 3 related posts

• Meta's Context Language Models: models that manage their own context hit +11.4% accuracy with 21.5% fewer FLOPs — meta · 2026-09-30
• Meta's Context Language Models treat context as an editable file, +65% on agent tasks — arankomatsuzaki · 2026-09-30

1 near-duplicate retellings: arankomatsuzaki