Using computers to read the entrails of modernity (statistics, optimization, machine learning). Currently: Stats PhD @Harvard Previously: CS/Soc @Stanford, Stat/ML @Oxford https://njw.fishhttps://bsky.app/profile/njw.fishhttps://bsky.app/profile/njw.fish/post/3mgn56gqfts22There's a growing worry that AI will break empirical social science -- that agents can p-hack until they find something that "works." We think that worry deserves to be taken seriously. Our new paper shows that is true empirically and makes it precise: https://njw.fish/static/papers/agentic_specification.pdf09 Mar 2026 15:09 +0000at://did:plc:d35hff3nuev7fqvo3tow6t6d/app.bsky.feed.post/3mgn56gqfts22https://bsky.app/profile/njw.fish/post/3mgflxi6urc2qNew paper: "Distribution-Conditioned Transport" Modern scientific datasets don't contain one population — they contain thousands. Clones, donors, patients, each its own distribution. DCT learns transport maps that generalize across them, including to distributions never seen during training.06 Mar 2026 15:12 +0000at://did:plc:d35hff3nuev7fqvo3tow6t6d/app.bsky.feed.post/3mgflxi6urc2qhttps://bsky.app/profile/njw.fish/post/3m76zgtk4mc2tThis project was accepted at #NeurIPS2025! We will be presenting today at 4:30 in Exhibit Hall C,D,E Poster Location: #1805! Swing by and learn about multiscale modeling, distribution representation learning, and some cool statistics! [contains quote post or other embedded content]04 Dec 2025 22:28 +0000at://did:plc:d35hff3nuev7fqvo3tow6t6d/app.bsky.feed.post/3m76zgtk4mc2thttps://bsky.app/profile/njw.fish/post/3lq3jm2wxic2z🚨 New preprint 🚨 We introduce Generative Distribution Embeddings (GDEs) — a framework for learning representations of distributions, not just datapoints. GDEs enable multiscale modeling and come with elegant statistical theory and some miraculous geometric results! 🧵26 May 2025 15:49 +0000at://did:plc:d35hff3nuev7fqvo3tow6t6d/app.bsky.feed.post/3lq3jm2wxic2zhttps://bsky.app/profile/njw.fish/post/3ld4qyxdiis2kExcited to share our new work on causal sensitivity analysis for fairness metrics at #NeurIPS2024! We've developed a causal sensitivity analysis framework to understand how underlying measurement biases (encoded by DAGs) impact machine learning fairness evaluations. 1 / 512 Dec 2024 17:06 +0000at://did:plc:d35hff3nuev7fqvo3tow6t6d/app.bsky.feed.post/3ld4qyxdiis2khttps://bsky.app/profile/njw.fish/post/3lcybarokns2pI’m at NeurIPS! If you’re in Vancouver and want to catch up let me know! I’ve been working at the intersection of generative models and causal inference, focused on bio/healthcare. If anything from active learning to causal foundation models sound interesting I’d love to chat!10 Dec 2024 22:14 +0000at://did:plc:d35hff3nuev7fqvo3tow6t6d/app.bsky.feed.post/3lcybarokns2phttps://bsky.app/profile/njw.fish/post/3k7tznnw4mx2qi guess its time to start posting here more regularly?20 Sep 2023 20:05 +0000at://did:plc:d35hff3nuev7fqvo3tow6t6d/app.bsky.feed.post/3k7tznnw4mx2qhttps://bsky.app/profile/njw.fish/post/3jzmyjxutaq2qanother day writing a custom qp solver, another day doing more linear algebra than could possibly be healthy03 Jul 2023 16:26 +0000at://did:plc:d35hff3nuev7fqvo3tow6t6d/app.bsky.feed.post/3jzmyjxutaq2qhttps://bsky.app/profile/njw.fish/post/3jzkvddfxjh2kare there any active set methods for least squares that allow for equality and inequality constraints while preserving sparsity? i have not found any...02 Jul 2023 20:23 +0000at://did:plc:d35hff3nuev7fqvo3tow6t6d/app.bsky.feed.post/3jzkvddfxjh2k