EpiEvolve: Self-Evolving Agents for Streaming Pandemic Forecasting under Regime Shifts

In three linesEpiEvolve is a self-evolving agent that adapts an LLM-based epidemic forecaster in streaming without weight updates. Using hierarchical episodic memory and reflection on errors, it achieves 0.629 accuracy on COVID-19 hospitalizations (vs 0.561 for static model), reducing recovery lag after regime shifts from 5 to 2 weeks.

Summary generated by Claude — human-verified