Agent harnesses enable recursive self-improvement without model weight updates
Evolved agents gain new tasks but lose covered ones, with reliability gains often raising deployment costs
Stories per day · 30 days
Relatedllm-agents2recursive-self-improvement2benchmark-evaluation1benchmarks1evaluation1forecasting1futurex1mldistillation1multi-view-distillation1open-source1reinforcement-learning1skill-memory1
4 stories, newest first · one card per story
Stories tagged #agent-evolution
MDSKILL internalizes skill memories via multi-view distillation for RL agents
MDSKILL improves ALFWorld and WebShop performance by up to 8.1% without external teachers.
RSIHub offers shared infrastructure for agent-evolution methods
The largest reported overall pass@1 gain is 13.5 points, but rankings and transfer vary by setting.
DeepEvolve agents self-improve by forecasting live events
Authors say the forecaster led FutureX for three weeks, with rating gains up to 195%.