[Discussion] Connecting factor stock pools with AI / RL trading research

[mykernel]

Hi FinRL team,

Thanks for maintaining one of the most important open-source projects for financial reinforcement learning. The data, environment, agent, and trading workflow in FinRL has been very helpful for people exploring AI in quantitative finance.

I am building a related but more research-workstation-oriented project: Factor Lab.

Factor Lab focuses on China A-share factor research:

• factor library and factor scoring;
• profit-gap / event-driven stock pools;
• AI-generated roundtable research reports;
• historical stock pool review and strategy replay.

Compared with FinRL, Factor Lab does not focus on reinforcement learning agents directly. It focuses more on candidate generation and research explanation before a trading model is trained or deployed.

Would love to exchange ideas on:

• whether factor-generated stock pools can serve as upstream candidate universes for RL models;
• how to evaluate event-driven signals such as profit-gap candidates;
• whether AI research reports can help explain or debug model decisions;
• how to bridge human-readable research workflows and automated trading agents.

Project link: https://www.afactorlab.com/