Ameya Daigavane

I am a PhD student at MIT EECS in Professor Tess Smidt's group, the Atomic Architects. I am broadly interested in the applications of machine learning to the computational sciences, especially at the intersection of equivariant models and generative modelling for biochemical systems. I am funded by the NSF Graduate Research Fellowship.

In 2024, I was an intern at Prescient Design where I worked on conformational ensembles of small peptides with Joseph Kleinhenz and Bodhi P. Vani.

Previously, I was a Pre-Doctoral Researcher at Google Research India, working with Prateek Jain and Gaurav Aggarwal on privacy-preserving graph neural networks.

I was a Caltech SURF Award recipient in 2019 and won the ACM SIGBED Scholars Award in 2020. At the time, I was an intern with the MLIA group at NASA Jet Propulsion Laboratory, where I worked with Kiri Wagstaff and Gary Doran on time-series algorithms to improve the responsiveness of the Plasma Instrument for Magnetic Sounding on the upcoming Europa Clipper mission.

In 2024, I received an S.M. in Electrical Engineering and Computer Science from MIT, where I was advised by Tess Smidt. I graduated with a B.Tech in Computer Science and Engineering (with a minor in Mathematics) from the Indian Institute of Technology, Guwahati in 2020.

I've also spent time as a volunteer at Youth4Jobs Shine where I mentored differently-abled youth on career planning and seeking employment, at English On Call where I taught English to non-native speakers, and at the MIT EECS GAAP program for helping underrepresented applicants with graduate school admissions.

EquiJump: Protein Dynamics Simulation via SO(3)-Equivariant Stochastic Interpolants
Allan dos Santos Costa, Ilan Mitnikov, Franco Pellegrini, Ameya Daigavane, Mario Geiger, Zhonglin Cao, Karsten Kreis, Tess Smidt, Emine Kucukbenli and Joseph Jacobson
GEM at ICLR'25
article | openreview

JAMUN: Transferable Molecular Conformational Ensemble Generation with Walk-Jump Sampling
Ameya Daigavane^*, Bodhi P. Vani^*, Saeed Saremi, Joshua Rackers and Joseph Kleinhenz
AIDrugX at NeurIPS'24
article | openreview | code

The Price of Freedom: Exploring Tradeoffs between Expressivity and Computational Efficiency in Equivariant Tensor Products
YuQing Xie, Ameya Daigavane, Mit Kotak and Tess Smidt
GRaM at ICML'24
openreview | code

Symphony: Symmetry-Equivariant Point-Centered Spherical Harmonics for Molecule Generation
Ameya Daigavane, Song Kim, Mario Geiger and Tess Smidt
ICLR'24
article | openreview | code

Differentially Private Graph Neural Networks for Medical Population Graphs and the Impact of the Graph Structure
Tamara T. Mueller, Maulik Chevli, Ameya Daigavane, Daniel Rueckert, and Georgios Kaissis
IEEE ISBI'24
article

Learning Integrable Dynamics with Action-Angle Networks
Ameya Daigavane, Arthur Kosmala, Miles Cranmer, Tess Smidt, and Shirley Ho
ML4PS at NeurIPS'22
article | poster | code

Unsupervised Detection of Saturn Magnetic Field Boundary Crossings from Plasma Spectrometer Data
Ameya Daigavane, Kiri Wagstaff, Gary Doran, Corey Cochrane, Caitriona Jackman, and Abigail Rymer
Computers and Geosciences
article | code

Resource Consumption and Radiation Tolerance Assessment for Data Analysis Algorithms Onboard Spacecraft
Gary Doran, Ameya Daigavane, and Kiri Wagstaff
IEEE Transactions on Aerospace and Electronic Systems
article | code

Integrating Deep Learning and Unbiased Automated High-Content Screening to Identify Complex Disease Signatures in Human Fibroblasts
Lauren Schiff, Bianca Migliori, et al
Nature Communications
article

Node-Level Differentially Private Graph Neural Networks
Ameya Daigavane, Gagan Madan, Aditya Sinha, Abhradeep Thakurta, Gaurav Aggarwal, and Prateek Jain
PAIR2Struct at ICLR'22
article | video | openreview

Understanding Convolutions on Graphs
Ameya Daigavane, Balaraman Ravindran, and Gaurav Aggarwal
Distill
article | code | reviews

Interactive Media for Understanding ML Methods: A Case-Study on Graph Neural Networks
Ameya Daigavane, Balaraman Ravindran and Gaurav Aggarwal
Rethinking ML Papers at ICLR'21
article | video | slides | openreview

Detection of Environment Transitions in Time Series Data for Responsive Science
Ameya Daigavane, Kiri Wagstaff, Gary Doran, Corey Cochrane, Caitriona Jackman and Abigail Rymer
MiLeTS at KDD'20
article | video | slides

Time-Series Analysis Methods for Onboard Detection of Magnetic Field Boundaries by Europa Clipper
Ameya Daigavane, Kiri Wagstaff, Gary Doran, Corey Cochrane, Caitriona Jackman and Abigail Rymer
Second AI and Data Science Workshop for Earth and Space Sciences
poster

2-uniform words: cycle graphs, and an algorithm to verify specific word-representations of graphs
Ameya Daigavane, Benny George and Mrityunjay Singh
Workshop on Words and Complexity
article | slides

An Introduction to (Modern) TensorFlow
Simran Khanuja and Ameya Daigavane
CVIT Summer School, IIIT Hyderabad
slides

Magnetic Field Boundaries in Cassini Plasma Spectrometer Data
Caitriona Jackman, Michelle Thomson, Michele Dougherty and Ameya Daigavane
An open-access dataset of CAPS ELS instrument observations with labelled crossing events.
zenodo | cite

Predicting Biological Activities of Molecules with Graph Neural Networks
Ameya Daigavane and Thomas Kipf
An official Flax example.
demo | source

Visualizing Graph Algorithms
Interactive visualizations of the incremental Delaunay triangulation and other graph algorithms.
demo | source