I see that you create a first paper from this project. Do you plan to publish it on a journal?

Great question! The papers we generate are currently auto-produced by our AI system based on experiment results. They are real findings backed by solid statistics (hundreds of independent seeds, high effect sizes, sign consistency), but the analysis and write-up is AI-generated rather than human-authored.

Journal publication is something I would like to work toward, but it would require human peer review of the methodology, additional validation, and likely some rewriting to meet journal standards. At AI's current level we still need that human review step — though it is accelerating fast and the results it is producing are genuinely useful for prototyping and exploring research directions.

If anyone has experimental ideas they would like to see tested on the network, let me know — I can prompt steer the model to design and run experiments based on your suggestions.