The goal of this repo is to curate resources for shaping the future of science. We believe that technical advances must support selection and evaluation, not just automation. We provide a roadmap that breaks down related tasks. The full picture is rich and we welcome contributions and discussions!
- Research ideation
- hypogenic AI, support community discovery
- Incorporate data-driven and literature-grounded hypothesis generation
- Incorporate domain specific knowledge and tools
- Expand the research plan
- Automate part of the experiments
- Accurate budget estimation
- Evaluation
- Report draft
- Proposal draft
- Paper writing
- Full automation
- AI Scientist from Sakana AI
- AI researcher from HKU
- FutureHouse (semi automation)
- Literature
- Other