Stars
Forks
Watchers
Developer links
AI Researcher
Accepted as a NeurIPS 2025 Spotlight paper and rapidly approaching 6,000 GitHub stars, AI-Researcher from the Hong Kong University Data Science Lab delivers the first fully autonomous scientific research system — a multi-agent platform that takes a list of reference papers and returns a complete research contribution with working code, validated experiments, and a formatted academic manuscript. The pipeline orchestrates five distinct phases: a Resource Collector systematically gathers materials from arXiv, IEEE Xplore, ACM Digital Library, Google Scholar, GitHub, and Hugging Face; an Idea Generator performs gap analysis against semantic embeddings to produce 3-5 novel hypotheses with feasibility scores; an Algorithm Designer transforms concepts into functional implementations; a Validation Engine automates testing, performance evaluation, and iterative optimization; and a Manuscript Creator generates polished full-length papers with figures, tables, and citations. The Gradio-based web GUI provides intuitive tabs for environment configuration, example selection, and real-time monitoring of research progress, while the production deployment at novix.science offers immediate browser access without local setup. Scientist-Bench provides a standardized benchmark comprising state-of-the-art papers across diverse AI research domains for evaluating autonomous research capabilities. The system supports multiple LLM providers including OpenAI, Anthropic, Google Gemini, and OpenRouter models with per-task routing for cost optimization. Deploy via Python with pip dependencies or Docker containerization. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. Apache-2.0 licensed.
Benefits
- End to End Research Automation
- Automates the complete scientific workflow from literature review through hypothesis formation, algorithm design, code implementation, experimental validation, and publication-ready manuscript generation.
- NeurIPS 2025 Spotlight Recognition
- Peer-reviewed and accepted as a Spotlight paper at one of machine learning's most prestigious conferences, validating the system's scientific rigor and practical utility.
- Multi Agent Architecture Design
- Specialized agents handle distinct research phases — resource collection, idea generation, algorithm design, validation, and writing — with coordinated handoffs between stages.
- Scientist Bench Evaluation Framework
- Comprehensive benchmark suite comprising state-of-the-art papers across diverse AI research domains enables standardized and reproducible evaluation of autonomous research system performance and quality.
Features
- Gradio Web Interface
- Browser-based GUI with tabs for environment configuration, example selection, research monitoring, and result inspection without command-line interaction.
- Academic Database Integration
- Resource Collector queries arXiv, IEEE Xplore, ACM Digital Library, Google Scholar, GitHub, and Hugging Face for comprehensive literature and code gathering.
- Multi Provider LLM Support
- Routes tasks to OpenAI, Anthropic, Google Gemini, or OpenRouter models with configurable per-phase model selection for cost and quality optimization.
- Iterative Experiment Validation
- Automated testing pipeline runs experiments against public benchmarks, evaluates performance metrics, and triggers iterative refinement cycles until convergence.
- Manuscript Generation Engine
- Produces publication-ready academic papers with structured sections, figures, tables, citations, and LaTeX formatting matching conference submission standards.