Science in the Age of Foundation Models
Published:
Featured perspective piece on how foundation models can be beneficial in scientific domains and their current limitations.
Danielle Maddix Robinson
Senior Applied Scientist, AWS AI Labs, PhD in Computational and Mathematical Engineering, Stanford University
- Bay Area, California
- Github
- Google Scholar
You May Also Enjoy
Mitra: Mixed synthetic priors for enhancing tabular foundation models
less than 1 minute read
Published:
Featured blog by Amazon Science on our tabular foundation model Mitra, which is trained purely on synthetic data and obtains state-of-the-art performance on classification and regression tasks. Mitra is available in AutoGluon 1.4, and we have also released the weights on HuggingFace. Find the Mitra classifier here and regressor here.
Physics-constrained machine learning for scientific computing
less than 1 minute read
Published:
Featured blog by Amazon Science on the research efforts that I am leading on physics-constrained machine learning for scientific computing and computational sciences.
A Lasting Legacy in Math and Science
less than 1 minute read
Published:
Blog, see pages 6-7 on how my parents inspired my love of math and science.
How applied math impacts forecasting at Amazon
less than 1 minute read
Published:
Featured blog by Amazon Science on my research on our award-winning paper on COVID-19 forecasting, where I advised our summer PhD student intern Rui (Ray) Wang.