AutoSci: A Memory-Centric Agentic System for the Full Scientific Research Lifecycle

2026-05-29

AI Summary

Traditional scientific research heavily relies on manual effort and lacks automated systems that span the entire research lifecycle while possessing structured, persistent memory and self-evolution capabilities. This work proposes a memory-centric research agent framework that innovatively decouples structured long-term memory from project-level active memory and integrates four core components: schema-constrained memory management (SciMem), a five-stage research workflow engine (SciFlow), a DAG-based multi-agent collaboration mechanism (SciDAG), and a feedback-driven evolution module (SciEvolve). The system enables end-to-end automation—from literature comprehension to peer-review response—while supporting cross-project memory reuse and continuous capability refinement, thereby establishing a unified, executable, memorable, and evolvable environment for automated scientific research.

Abstract

Scientific research has traditionally been human-intensive, requiring researchers to coordinate literature, ideas, experiments, manuscripts, and review responses across long project cycles. The rise of LLM-based scientific agents creates an opportunity to automate this process. Such a system must support the full research lifecycle, maintain structured persistent memory across projects, and improve its own research procedures over time. However, existing systems either partially satisfy or fail to satisfy these requirements, leaving a gap for a unified automated scientific research system. As a result, we present AutoSci, a memory-centric agentic system for the full scientific research lifecycle. AutoSci is organized around four modules. SciMem provides schema-governed research memory, separating Long-Term Knowledge Memory for reusable scientific knowledge from Active Research Memory for project-level artifacts such as ideas, experiments, manuscripts, and reviews. SciFlow executes a five-stage lifecycle from literature understanding to rebuttal through a harness that controls state, context, verification, feedback, and orchestration. SciDAG augments difficult skills with DAG-shaped multi-agent operators and reusable stage-specific templates. SciEvolve converts feedback signals from users, experiments, reviews, and external environments into versioned updates to SciMem organization, SciFlow skills, and SciDAG templates. Together, these modules make AutoSci a persistent research environment that can execute, remember, and evolve across research projects. The code repository is available at https://github.com/skyllwt/AutoSci.

Problem

• scientific research automation
• persistent memory
• research lifecycle
• LLM-based agents
• automated scientific system

Innovation

• memory-centric architecture
• scientific agent
• research lifecycle automation
• persistent memory
• self-evolving system