Research Scientist Manager, Adaptive Experimentation team @Meta. Creator of BoTorch. Interested in Bayesian Optimization and the like. Scuba Instructor.
Proud to share the 1.0 release of Ax, our platform for adaptive experimentation platform. We've been using it for the past 6 years for optimizing everything from end-to-end AI systems, recommender systems, AR hardware, and material science applications.
https://t.co/19aLoJ6jJU
Excited to share our ICML 2025 paper:
"Scalable Gaussian Processes with Latent Kronecker Structure"
We unlock efficient linear algebra for your kernel matrix which *almost* has Kronecker product structure.
Check out our paper here:
https://t.co/wqq89CTrAb
Compute is increasing much faster than data. How can we improve classical supervised learning long term (the underlying tech of most of GenAI)?
Our ICML position paper's answer: simply train on a bunch of artificial data (noise) and only do inference on real-world data! 1/n
2024 PhD Internship opportunity! Join me and my amazing colleagues on Meta's Adaptive Experimentation team to work on Bayesian optimization, probabilistic modeling / Gaussian Processes, and sample-efficient decision making: https://t.co/5uR0NNXLQU
Heading to NOLA for #NeurIPS2023 - find me at our (spotlight!) poster on logEI (Poster Session 2 on Tue): https://t.co/56iWiFpTdz - or DM me if you want to chat about Bayesian Optimization on the side.
P.S.: We're hiring PhD Research Interns for 2024: https://t.co/VTjDPE8iU8
2024 PhD Internship opportunity! Join me and my amazing colleagues on Meta's Adaptive Experimentation team to work on Bayesian optimization, probabilistic modeling / Gaussian Processes, and sample-efficient decision making: https://t.co/5uR0NNXLQU
Pleased to share that we have a post-doctoral position on the Adaptive Experimentation team at Meta. Please share this on with others interested in Bayesian optimization, active learning, or Bayesian deep learning! https://t.co/wmvyx6Q5zU
@JPDuerholt @felix_s_k @procrastiprof @leecronin@A_Aspuru_Guzik@felix_s_k What exactly are the shortcomings / issues you observed for batch acquisition functions and discrete spaces in BoTorch? Sequential greedy + conditioning using pending should still work? Or is the issue that there are too many pending points?
High-Dimensional Combinatorial BO via Dictionary-based Embeddings (BODi) will be presented at AISTATS @aistats_conf next week.
Paper: https://t.co/JJxlkTqGRn
Code: https://t.co/7N8jPs7T1B
w/ @deshwal_aryan, @SebastianAment, @davidmeriksson, @MaxBalandat, and @eytan
1/N
Are you attending #SIAMCSE23 and want to learn about how Bayesian optimization can be used for problems such as peptide design, molecular design, and optimizing stellarator coils?
Join @MaxBalandat and I tomorrow in our minisymposium on Bayesian optimization in the real-world.
@JPDuerholt @eytan Indeed! They’ve got fancier animations for their twitter posts though :) We do have an internal version of Ax as an RPC service as well; I wonder if there would be value in open sourcing this as well at some point.
On this darkest day of the year, allow me to share a bit of light in the form of a fantastic tutorial on adaptive experimentation put together by my student @SterlingBaird1. Here is the first video in the series: https://t.co/PATRPZdiRM
new posts through Christmas ;)
I’m looking forward to presenting our poster on LinearOperator [https://t.co/7NDWOssKpc] - a library for structured linear algebra in @PyTorch - at this year's #PyTorchConference (co-located w/ NeurIPS).
If you want to meet @MaxBalandat & @davidmeriksson and learn more, register here for #PyTorchConference on December 2:
🤝In-Person in New Orleans Registration: https://t.co/9awTFMcpus
📺Virtual Streaming Event RSVP: https://t.co/ZbZN8BTi0o
4/4