Hugging Face launches Paper Reproductions with ML Intern for agent-run replication

Hugging Face has launched Paper Reproductions with ML Intern, a feature that lets users task agents with independently reproducing the experiments of a published paper.

The company says the goal is to generate more signals, such as open artifacts, that help identify high-quality research. The announcement came via its official X account, which framed the feature as part of Paper Pages on hf.co/papers. The reporting available here is a single launch post, so it does not document results, accuracy of reproductions, or how widely the feature is available.