Meta's Self-Improving Memory: model meta-optimizes its own memory prompts, then distills gains into weights
jaseweston · x · 2026-11-17
Jason Weston (Meta AI) introduces Self-Improving Memory: the model discovers its own memory skills from its runs, then trains them into its weights.
- Meta-optimization of memory from the model's own runs yields large gains;
- Distilling these gains into weights works too.
Background: different tasks empirically need different memory methods. Baseline: each new task starts with empty context plus a carried-over memory system the agent reads/writes. The team hand-wrote 4 memory prompt designs, and the best one changes per task.
Related event: Meta's Self-Improving Memory Lets Agents Discover and Distill Memory Skills(4 posts)→
More from coding & agent
- How Comet improved a financial incident remediation agent's score from 0.34 to 0.57 with Opik — dl_weekly · 2026-11-17
- Jason Weston: Harness Optimization Hits 56.0, Distillation Adds 5.5 More — jaseweston · 2026-11-17
- Meta's Self-Improving Memory: agents discover their own memory skills and distill them into weights — jaseweston · 2026-11-17
- Every shows Claude Fable 5.2 building interactive games — slow but token-hungry — every · 2026-11-17
- Two free Wan2.1 ComfyUI workflow sets: native nodes vs Kijai wrapper with speed boosts — asdfghmfker · 2026-11-17
- Keeping filler words in prompts encodes real uncertainty and nails it in one shot — brandon_galang · 2026-11-17