Pick a result from a NeurIPS 2026 paper. Try to reproduce it with ML Intern. Share your code, logs and verdict so anyone can check your work.
A claim is a result that a paper reports, for example “our model gets 94.1% accuracy (Table 2)”. In this hackathon you:
Your verdict for each claim is one of: verified (you got the same result), falsified (the claim does not hold), or inconclusive (you could not tell).
Reproduce claims from NeurIPS 2026 papers and report your verdict. This is the main track.
Turn a reproduction into an environment that AI agents can practise on, using the Harbor or OpenEnv format. Details TBD
Build something new from a reproduction: a visualization that explains the paper, or a Gradio workflow app that uses its method.
Click “Request to join this org” at the top of this page. Then fill in the registration form TBD. Registration closes the day before the hackathon starts.
Use our claim list TBD, or pick claims from any other NeurIPS 2026 paper.
Run the experiments. Keep your code, outputs and logs.
Put your work in a public Space and send it through the submission form. You can edit and resubmit until the deadline.
We are preparing a list of claims from NeurIPS 2026 papers that have released code or model weights. Each claim has an ID (for example neurips26-4821/c2) and points to a figure, table or theorem in the paper.
Release date of the list: TBD
For each claim, your Space must contain:
| Item | What it means |
|---|---|
| Code | One script that runs the experiment from start to finish. |
| Environment | The exact software versions you used (for example requirements.txt, uv.lock or a Dockerfile). |
| Results | The number the paper reports and the number you got. |
| Logs | The full output of each run, including runs that failed. |
| Verdict | Verified, falsified or inconclusive, with a short reason. |
| Agent trace | The log of your AI agent session, saved to a Hub dataset. ML Intern will have a “Publish Trace” button for this Coming soon |
Then fill in the submission form. It is a Gradio app that checks your Space, shows a checklist of what is complete and what is missing, and marks your submission as complete. AI agents can submit through the same form, because it is also an MCP server. The form also asks which agent and model you used, how much you helped the agent, and whether you looked at other people's reproductions.
Full judging criteria: TBD
Careful, complete and easy to re-run.
A claim that does not hold, with a script that proves it and a clear explanation of the problem.
The clearest visual explanation of a paper's key idea.
The most useful Gradio workflow app built on a paper.
The entry the community likes most.
TBD
Prize amounts: TBD
For every registered participant. How and when you receive them: TBD
Hugging Face's AI agent for ML work. It can read papers, write code, run jobs and save results to the Hub. Try ML Intern
A Gradio app and MCP server that checks your Space and records your submission. Link: TBD
| Date | What happens |
|---|---|
| TBD | Registration opens |
| Nov 19 Tentative | Registration closes |
| Nov 20 Tentative | Hackathon starts |
| Dec 5 Tentative | Submission deadline |
| TBD | Winners announced |
NeurIPS 2026 takes place December 6–13. Plans to share results there: TBD
We will publish the claim list, all submissions, agent traces and verdicts as open datasets on the Hub. Anyone will be able to see which NeurIPS 2026 results were reproduced.
No. You can use any agent, for example Claude Code, Codex or Cursor. Say what you used in the submission form. ML Intern is the main focus of the hackathon.
Submit it anyway. A careful “falsified” or “inconclusive” verdict is a useful result.
We want careful work on a few papers, not quick work on many.
Yes, as many times as you like until the deadline.
No. Registration closes the day before the start.
All submissions are public and open-sourced, and may be used to build open datasets and benchmarks.
Questions: Discord channel TBD · @Gradio on X
By taking part you agree to the hackathon terms and conditions TBD, the Hugging Face Code of Conduct and the Content Policy.