tongjingqi/AI-Can-Learn-Scientific-Taste
A framework for training AI agents to evaluate and propose high-impact scientific research ideas using community feedback signals.
Not currently ranked — collecting fresh signals.
star history
The project proposes Reinforcement Learning from Community Feedback (RLCF), a training paradigm that uses large-scale community signals as supervision to teach AI systems scientific taste. It formulates scientific taste learning as a preference modeling and alignment problem, enabling AI to judge and propose research ideas with high potential impact. The framework releases models and datasets on Hugging Face and includes an online demo for paper evaluation.
Frequently asked
- What is tongjingqi/AI-Can-Learn-Scientific-Taste?
- A framework for training AI agents to evaluate and propose high-impact scientific research ideas using community feedback signals.
- Is AI-Can-Learn-Scientific-Taste open source?
- Yes — tongjingqi/AI-Can-Learn-Scientific-Taste is open source, released under the Apache-2.0 license.
- How popular is AI-Can-Learn-Scientific-Taste?
- tongjingqi/AI-Can-Learn-Scientific-Taste has 418 stars on GitHub.
- Where can I find AI-Can-Learn-Scientific-Taste?
- tongjingqi/AI-Can-Learn-Scientific-Taste is on GitHub at https://github.com/tongjingqi/AI-Can-Learn-Scientific-Taste.