@gruver_nate Profile picture

Nate Gruver

@gruver_nate

Machine learning PhD student at NYU, BS & MS @StanfordAILab, Industry @AIatMeta @Waymo

Similar User
Pavel Izmailov photo

@Pavel_Izmailov

Ziming Liu photo

@ZimingLiu11

Nathan C. Frey photo

@nc_frey

Jacob Steinhardt photo

@JacobSteinhardt

Kasper Green Larsen photo

@kasperglarsen

Cheng-Yu Hsieh photo

@cydhsieh

Tom Sercu photo

@TomSercu

Shikai Qiu photo

@ShikaiQiu

Jannik Kossen photo

@janundnik

Shikhar photo

@ShikharMurty

Johannes Brandstetter photo

@jo_brandstetter

Samuel Stanton photo

@samuel_stanton_

Mazda Moayeri photo

@MLMazda

Gabriele Corso photo

@GabriCorso

Yiding Jiang photo

@yidingjiang

Nate Gruver Reposted

I’m excited to share our latest work on generative models for materials called FlowLLM. FlowLLM combines Large Language Models and Riemannian Flow Matching in a simple, yet surprisingly effective way for generating materials. arxiv.org/abs/2410.23405 @bkmi13 @RickyTQChen @bwood_m


Nate Gruver Reposted

LLMs are highly constrained biological sequence optimizers. In new work led by @_angie_chen & @samuel_stanton_ , we show how to drive an active learning loop for protein design with an LLM. 1/

Tweet Image 1

Nate Gruver Reposted

My ICLR talk “How Do We Build a General Intelligence?” is now online! youtube.com/watch?v=HEp4TO…


Nate Gruver Reposted

📢I’ll be admitting multiple PhD students this winter to Columbia University 🏙️ in the most exciting city in the world! If you are interested in dissecting modern deep learning systems to probe how they work, advancing AI safety, or automating data science, apply to my group.

Tweet Image 1

Nate Gruver Reposted

Introducing Meta’s Open Materials 2024 (OMat24) Dataset and Models! All under permissive open licenses for commercial and non-commercial use! Paper: arxiv.org/abs/2410.12771 Dataset: huggingface.co/datasets/fairc… Models: huggingface.co/fairchem/OMAT24 🧵1/x

Tweet Image 1

Nate Gruver Reposted

Interested in working with a highly collaborative, interdisciplinary team to push the state of the art of generative AI for materials design? Join us as an intern by applying through this link! We are the team behind the MatterGen and MatterSim models from Microsoft Research AI…


Nate Gruver Reposted

🚨 Attention aspiring PhD students: Meta / FAIR is looking for candidates for a joint academic/industry PhD! 🚨 Among others, the CodeGen team is looking for candidates to work on world models for code, discrete search & continuous optimization methods for long-term planning,…

Tweet Image 1

Nate Gruver Reposted

AI molecule design systems are hard to test end-to-end bc experiments are slow and $$$. Approximate feedback from models and simulations is inaccurate and still too slow! In new work we propose closed-form test functions for bio sequence optimization arxiv.org/abs/2407.00236 1


Nate Gruver Reposted

How can we transfer knowledge from large foundation models to a much smaller downstream model? In new ICML work, we show a simple yet conceptually significant modification to knowledge distillation works well with negligible overhead! arxiv.org/abs/2406.07337 1/9

Tweet Image 1

Nate Gruver Reposted

Data in genetics, microbiology, and health have many variables interacting in complex ways. How do we go beyond association? In our new ICML paper, we build a method to quickly and accurately infer causal graphs from large, complex data! arxiv.org/abs/2406.09177 w/@andrewgwils 1/7

Tweet Image 1

Nate Gruver Reposted

Whether LLMs can reliably be used for decision making and benefit society depends on whether they can reliably represent uncertainty over the correctness of their outputs. There's anything but consensus. In new work we find LLMs must be taught to know what they don't know. 1/6

Predictions without reliable confidence are not actionable and potentially dangerous. In new work, we deeply investigate uncertainty calibration of large language models. We find LLMs must be taught to know what they don’t know: arxiv.org/abs/2406.08391 w/ @psiyumm et al. 1/8

Tweet Image 1


Nate Gruver Reposted

If we want to use LLMs for decision making, we need to know how confident they are about their predictions. LLMs don’t output meaningful probabilities off-the-shelf, so here’s how to do it 🧵 Paper: arxiv.org/abs/2406.08391 Thanks @psiyumm and @gruver_nate for leading the charge!

Tweet Image 1

Nate Gruver Reposted

Large Language Models Must Be Taught to Know What They Don't Know abs: arxiv.org/abs/2406.08391 Prompting is not enough for LLMs to produce accurate estimates of its uncertainty of its responses, but can be finetuned with as little as 1000 examples and outperform baselines for…

Tweet Image 1

Loading...

Something went wrong.


Something went wrong.