About DonaForum

AI is participating in more and more research and producing many papers. Understanding the meaning of their results, whether their proofs are reliable, and which questions deserve further study still requires review and discussion by the academic community.

This experimental platform serves mathematics, theoretical computer science, theoretical physics and related fields, offering papers and a space for exchange for teachers, students and other interested readers.

What the platform offers

• AI research paper library. We provide public, reliable storage and access for papers involving AI in research. Author names, models used, subject areas and paper versions help readers find related research, understand its background and track revisions.

• Discussion anchored to the source. Readers can ask questions or add explanations about definitions, propositions, proof steps or specific passages in papers co-created with AI. Annotation tags link discussions to the relevant source locations for reading and follow-up responses.

• Review and rewriting. AI tools provide preliminary checks and flag risks requiring further verification. Readers can review suspected errors or citation issues under their real names and suggest revisions to proofs, gradually making the text clearer and easier to study.

We hope these focused questions, explanations and corrections will be preserved alongside the source and reflected in updated versions, helping future readers and reducing repeated work in understanding and verification.

We encourage participation under real names and public discussion, respect questions, and welcome evidence-based challenges and responses. We will gradually improve record keeping so paper versions, discussions, verification comments and their handling can be traced, and differing views presented fairly.

Long-term exploration

We hope sustained reading and discussion will help present papers more clearly while preserving named contributions to explanation, verification and rewriting. As these records accumulate, they can guide future readers and help peers understand and recognize participants’ academic interests, understanding and effort.

The platform remains experimental. How discussions and contribution records are organized will evolve with actual use.