Meta FAIR's Self-Improving Memory lets agents rewrite their own memory prompts
Start here
Meta FAIR, with researchers from the University of Washington and Johns Hopkins University, introduced Self-Improving Memory, a two-stage method that lets a Kimi-K2.6 agent rewrite its own LOAD and OFFLOAD memory prompts and then distill the resulting skills into LoRA weights. The optimized harness averaged 56.0 across the Database, Codebase and Spectrum tasks of Continual Learning Bench, versus 44.6 for the best human-written prompts, with Codebase gaining 20.6 points to 64.3. After distillation the model runs the generic prompts and still averages 43.1, against 37.6 for SFT on generic trajectories. The researchers report that 37 of 44 meta-agent revisions failed to beat the previous best.
Generated from reports · updated 27 min ago
Report timeline
Follow the reports to see the event from different sides.
- MarkTechPostMeta FAIR's Self-Improving Memory scores 56.0 on Continual Learning Bench
Meta FAIR, with researchers from the University of Washington and Johns Hopkins University, introduced Self-Improving Memory, a two-stage method that lets a Kimi-K2.6 agent rewrite its own LOAD and OFFLOAD memory prompts and then distill the resulting skills into LoRA weights. The optimized harness averaged 56.0 across the Database, Codebase and Spectrum tasks of Continual Learning Bench, versus 44.6 for the best human-written prompts, with Codebase gaining 20.6 points to 64.3. After distillation the model runs the generic prompts and still averages 43.1, against 37.6 for SFT on generic trajectories, though 37 of 44 meta-agent revisions failed to beat the previous best.
Heat for this event
Not enough continuous observations yet to draw a trend.