Profile avatar
nmboffi.bsky.social
Building generative AI systems for high-dimensional science and engineering. Assistant prof. @CarnegieMellon & @mldcmu, PhD @Harvard / @MIT, postdoc @NYU_Courant. https://nmboffi.github.io
33 posts 155 followers 496 following
Getting Started
Active Commenter
Posts 9 Comments 26

this paper is a pretty impressive tour de force in neural network training: arxiv.org/abs/2410.11081 pretty inspiring to me -- network isn't converging? rigorously monitor every term in your loss to identify where in the architecture something is going wrong!

submitted 15 days ago • 0 comments

Oscillations naturally emerge in large crowds. That's because your brain works by oscillating. Emergence of collective oscillations in massive human crowds www.nature.com/articles/s41... #neuroscience

submitted 18 days ago • 3 comments

what's the status on using jax for image experiments (e.g. with diffusion models)? it seems like standard packages like huggingface diffusers have much less robust implementations of the same neural networks than the pytorch counterpart?

submitted 25 days ago • 2 comments

📣 Excited to receive the 💥 NSF CAREER 💥 Award. Our group is looking for PhD students and postdocs interested in the Nonlinear Dynamics and the Physics of Living Systems, with support from NSF, HFSP, and other sources. We’d appreciate your help in spreading the word! @ucsdphysci.bsky.social

submitted 43 days ago • 1 comment

some really nice new work by @jiequnh.bsky.social and collaborators arxiv.org/abs/2409.08526 of course i'm super biased, but i think that figuring out how to solve high-dimensional scientific computing problems with ML has potentially very high impact. i'd love to see more work like this

submitted 43 days ago • 1 comment

what are the most important open problems in molecular simulation right now that stand to benefit from ML-based methods? any good reviews or references to get up to speed rapidly? @gcorso.bsky.social @hannes-stark.bsky.social?

submitted 49 days ago • 1 comment

there's a lot of interest right now in using diffusion and flow matching models for sampling (i.e., no data but access to the energy). is anyone aware of works using diffusion models or score-based approaches for data assimilation? seems like it could be a natural evolution of kalman filter ideas.

submitted 52 days ago • 0 comments

question for the academics: how do you manage your (ever-growing) reading list? do you have a few dedicated reading blocks per week, where you tackle some papers you've been interested in? do you only read as you need to solve a research problem? how do you fit textbooks into this?

submitted 53 days ago • 2 comments

quite excited to read this in the new year: arxiv.org/abs/2410.10523

submitted 55 days ago • 0 comments