EvoLib: Turning Experience into Evolving Knowledge

Summary

Microsoft Research introduces EvoLib, a framework that enables large language models to learn from their own experience during inference without ground-truth labels or external feedback. EvoLib transforms past attempts into reusable skills and reflective insights, continually refining, consolidating, and reweighting them to turn instance-specific observations into increasingly general knowledge over time. This evolving knowledge transfers across tasks, helping AI models learn from successes and failures, and consistently improves downstream performance. Since EvoLib does not require model updates, it can be applied to any black-box language models and AI systems deployed through APIs.