New paper from my friends at Inherent (who love open-endedness like I do), where they train a new AI scientist model that's particularly good and efficient at experimental replication: arxiv.org/pdf/2608.13331
arxiv.org
arxiv.org
Jimmy Secretan
@jsecretan.bsky.social
398 followers 2.1K following 34 posts
Interested in AI, open endedness, privacy
