Algorithm discovery improves with self-evolving interactive search agent
AlgoEvo: Self-Evolving Agentic Search for Automated Algorithm Discovery
Artificial Intelligence
Summary
Designing new algorithms automatically is hard because existing methods follow fixed rules that don’t adapt well to what’s learned during the process. The authors propose AlgoEvo, a system where an intelligent agent tests, diagnoses, and improves its own code step-by-step using real-time feedback. It also organizes knowledge into reusable skills that help solve different but related problems more efficiently. Their experiments show AlgoEvo can match or beat other specialized methods while using fewer resources and learning across tasks.
What this means in practice
- •For software development teams: Automatically generate and improve algorithms for complex coding tasks by integrating runtime diagnostics into the development workflow.
- •For robotics engineers: Discover and refine control algorithms that adapt efficiently to multiple objectives in robotic systems using interactive agent feedback.
Authors
Junhao Qiu, Qinglong Hu, Xialiang Tong, Mingxuan Yuan, Liyong Lin, Qingfu Zhang
Abstract
Large language models have advanced automated algorithm discovery by synthesizing executable code, but existing frameworks trap them in rigid search pipelines with pre-defined control flows. This limitation restricts adaptive reasoning, blocks cross-paradigm transfer, and discards valuable execution feedback. We propose AlgoEvo, a unified agentic framework that transforms automated algorithm discovery into an interactive, knowledge-accumulating process. An autonomous agent dynamically inspects, diagnoses, and edits code based on runtime feedback. A design skill hub decouples paradigm-specific knowledge from the core discovery engine, allowing a single workflow to seamlessly handle single-objective, multi-objective, and multi-component design. Meanwhile, a hierarchical experience mechanism organizes search trajectories into a task-level tree to guide exploration and consolidates cross-task patterns into reusable skills. Across six representative benchmark tasks, AlgoEvo matches or surpasses specialized methods with substantially fewer evaluations and reduced token consumption, demonstrating strong intra-task accumulation, cross-task transfer, and the ability to reproduce or exceed existing state-of-the-art performance through flexible skill activation.