The Hitchhiker’s Guide to Autonomous Research: A Survey of Scientific Agents
Путеводитель автостопщика по автономным исследованиям: обзор научных агентов
2026-08-07
SCID: 54.1/gzthb6h9
Discuss with AI
AI for Science (AI4S)AWESOME_SCIENTIFIC_AGENTlarge language model agentsscientific agentsscientific research lifecycle
Figures from the paper
Abstract (AI)
The advancement of LLM-based agents is redefining AI for Science (AI4S) by enabling autonomous scientific research. Prominent LLMs exhibited expertise across multiple domains, catalysing constructions of domain-specialised scientific agents. Nevertheless, the profound epistemic and methodological gaps between AI and the natural sciences still impede the systematic design, training, and validation of these agents. This survey bridges the existing gap by presenting a comprehensive blueprint for scientific agents' design. It first clarifies the concept of scientific agents and distinguishes them from general-purpose agents in terms of their goal orientation, workflow embedding, and scientific commitments. It then introduces a unified taxonomy based on capability envelope and capability maturity, characterizing both the scope of scientific workflow coverage and the reliability of agent behavior under realistic research conditions. Building on this taxonomy, the survey further connects scientific agent design with the research life cycle by reviewing construction strategies, capability enhancement methods, evaluation paradigms, and future challenges. This unified perspective aims to provide practical guidance for designing domain-specific scientific agents and to promote the convergence of AI research and natural scientific discovery. To support long-term progress, we curate a live repository (https://github.com/gudehhh666/Awesome_Scientific_Agent Awesome Scientific Agent) that continuously aggregates emerging methods, benchmarks, and best practices.
Key Findings
1
A unified taxonomy characterizes scientific agents by capability envelope, covering workflow scope, and capability maturity, reflecting behavioral reliability under realistic research conditions.
2
Scientific agents are distinguished from general-purpose agents by goal orientation, integration into scientific workflows, and explicit scientific commitments.
3
The live AWESOME_SCIENTIFIC_AGENT repository continuously aggregates emerging methods, benchmarks, and best practices to support long-term progress.
4
The survey maps scientific-agent construction, capability enhancement, and evaluation strategies onto the full scientific research life cycle.
5
The survey provides a comprehensive blueprint for designing, training, evaluating, and advancing LLM-based scientific agents.
Research Object
LLM-based scientific agents for autonomous research
Research Subject
Their design, taxonomy, capabilities, reliability, evaluation, and integration into the scientific research life cycle
Publication Details
Publication Date
2026-08-07
Journal
Publisher
ISSN
Cited by
4
Open access PDF
Access Type
Author Information
Download PDF
Subscribe to digest
References available in scid.ai15
Highly accurate protein structure prediction with AlphaFold2021
The AI Scientist: Towards Fully Automated Open-Ended Scientific Discovery2024
Autonomous chemical research with large language models2023
An autonomous laboratory for the accelerated synthesis of inorganic materials2023
Augmenting large language models with chemistry tools2024
AgentClinic: a multimodal benchmark for tool-using clinical AI agents2026
Autonomous biomedical research with an artificial intelligence agent2026
Empowering biomedical discovery with AI agents2024
A multi-agent system for automating scientific discovery2026
AI Agents vs. Agentic AI: A Conceptual taxonomy, applications and challenges2025
ChatMOF: an artificial intelligence system for predicting and generating metal-organic frameworks using large language models2024
Accelerating scientific discovery with Co-Scientist2026
An agentic system for rare disease diagnosis with traceable reasoning2026
Biomni: A General-Purpose Biomedical AI Agent2025
AI, agentic models and lab automation for scientific discovery — the beginning of scAInce2025