SelfAI: A self-directed framework for long-horizon scientific discovery

TL;DR

SelfAI is a multi-agent system that automates long-horizon scientific discovery by strategically guiding experiments, balancing efficiency and diversity, and supporting human-in-the-loop workflows, demonstrated across various scientific domains.

Contribution

It introduces a novel self-directed, multi-agent framework for scientific exploration that optimizes long-term discovery trajectories with adaptive decision-making.

Findings

SelfAI discovers high-quality solutions with fewer trials than classical methods.

It effectively balances exploration diversity and search efficiency.

Demonstrated success across machine learning and drug discovery domains.

Abstract

Scientific discovery increasingly entails long-horizon exploration of complex hypothesis spaces, yet most existing approaches emphasize final performance while offering limited insight into how scientific exploration unfolds over time, particularly balancing efficiency-diversity trade-offs and supporting reproducible, human-in-the-loop discovery workflows. We introduce SelfAI, a self-directed, multi-agent-enabled discovery system that automates scientific exploration as a strategic, trajectory-driven decision-making process. SelfAI translates high-level research intent into executable experiments, reasons over accumulated experimental trajectories to guide subsequent exploration, and applies adaptive stopping decisions to terminate unproductive search paths within a closed-loop workflow governed by explicit efficiency-diversity trade-offs. Evaluated using real-world experiments spanning domains from machine learning to drug discovery, SelfAI consistently discovers high-quality solutions with substantially fewer redundant trials than classical optimization and recent LLM-based baselines. The proposed methods establish a general framework for organizing long-horizon scientific discovery and adaptive decision-making in complex scientific and engineering systems.