automated-science 开源项目推荐（15 个）

mims-harvard /ToolUniverse

用 ToolUniverse 普及 AI 科学家

mims-harvard /AutoScientists

AutoScientists：长期科学实验的自组织代理团队

Sibyl-Research-Team /AutoResearch-SibylSystem

具备自我进化能力的完全自主 AI 研究系统，原生构建于 Claude Code 之上

InternScience /SGI-Bench

Probing Scientific General Intelligence of LLMs with Scientist-Aligned Workflows

Omni-Scientist /Awesome-AI-Scientist

🧪 Awesome list of AI Scientist papers, systems, benchmarks, datasets and open-source platforms.

ssocean /NAIP

This repository contains the official implementation for the series of NAIP family.

cloudlab-cmu /Public-Protocols

The academic Cloud Lab is primarily for the use of the CMU research community and CMU educational purposes, providing access to a wide array of instrumentation. The Academic Cloud Lab has already been used to teach courses for both undergraduate and graduate students, and CMU researchers have conducted research using ECL facilities.

YUANXICHE98 /LabOS

🔬 LabOS — 下一代全自动科研实验平台 | Next-Gen AI Research Automation Platform | 四阶段流水线 · 多LLM配置 · Codex集成 · 远程GPU执行 · 记忆系统

Cosmos-Logic-Institute-CLI /Infinite-Alchemy

One Day’s Operation, One Year’s Discovery

docxology /template_autoscientists

Testbed for the five coordination mechanisms in the AutoScientists paper on self-organizing agent teams: shared experiment log, dead-end registry (36 redundant re-probes to 0), effect-size ranking, noise-band confirmation, stagnation reorganization. Python, uv, no mocks.

kadubon /audit-closed-ai-scientist

Benchmark for statistically valid AI scientist systems, using audit-closed protocols, transparency logs, and sequential inference to prevent false discoveries in autonomous research agents.

EnergeticParadigm /ERPL

Public taster of ERPL (EP Research Production Line) — a capability-limited, executable and independently verifiable preview of an AI-assisted research-production architecture spanning discovery, hypothesis, evidence, review, revision, and closure. The full private runtime is not included.

Garbologywhiteleavedrockrose661 /autoresearch

Automate code improvement by detecting issues, fixing bugs, and simplifying code on separate branches before merging to main.

Advanced-Research-Centre /HilbertCorps

Automated Concept Discovery of Quantum Error Correction Codes

cloudlab-cmu /.github

该项目在 GitHub 上没有填写描述