I am a researcher in computational materials science interested in the intersection of machine learning with traditional simulation techniques. My recent work has spanned computational materials discovery, machine learning applied to first principles simulation, and battery informatics.

I currently work as a senior research scientist at Toyota Research Institute. You can view the papers that I’ve written on my Google Scholar.

I am passionate about serving the scientific and academic communities in my spare time. I currently serve on the external advisory board of the UC Merced Department of Mechanical Engineering.

Prior to that, I was a DOE Computational Science Graduate Fellow, and a Barry Goldwater scholar. I obtained my Ph.D. in Physics with a secondary in computational science & engineering from Harvard University, where I worked under Prof. Boris Kozinsky, and a B.S. in Physics and a B.A. in Mathematics from the University of Rochester.

Last Updated: May 26, 2025

📢 Recent & On The Horizon

• Summer 25
  • It was my honor to serve on the dissertation committee of Dr. Tzu-chen Liu (a collaborator and student of Prof. Chris Wolverton) at Northwestern University.
  • I will be speaking at the NIST Artificial Intelligence for Materials Science workshop in July!
  • I will be speaking at the Telluride Science Research Center’s conference on Machine Learning for Chemistry and Materials Science for the third time this September!
• My work with Tzu-chen Liu, Prof. Chris Wolverton, and other students in his group on the influence of the U parameter on the stability of Mo-containing oxides has been published in Physical Review Materials!
• My work with Dr. Tina Na Narong, Zoe Zachko, and Prof. Simon Billinge on interpretable multimodal analysis of pair distribution functions and X-ray absorprion spectroscopy has been published in NPJ Computational Materials!

🧪 Research Directions

My work sits at the intersection of machine learning, computational materials science, and data-driven discovery. I focus on building interpretable and scalable models that interface with first-principles simulations, spectroscopic experiments, and electrochemical devices.

1. Machine Learning from First-Principles Simulations

I am interested in ML models — including surrogate models, interatomic potentials, and property predictors — that accelerate traditional quantum simulations and uncover patterns in atomistic simulation.

Selected Works:

• 📄 Simultaneous Discovery of Reaction Coordinates and Committor Functions Using Equivariant Graph Neural Networks
  Sheriff, Freitas, Trewartha, Torrisi, AI for Materials @ NeurIPS Workshop 2024
• 📄 Benchmarking FLARE and NequIP with the TM23 Dataset
  Owen & Torrisi, npj Computational Materials (Co-first author)
• 📄 On-the-Fly Active Learning of Bayesian Force Fields
  Vandermause, Torrisi, et al., npj Computational Materials

2. Interpretable Spectroscopy

I am interested ins ML frameworks to interpret spectroscopic data — particularly X-ray absorption and PDF — in a way that allows practitioners to build chemical intuition and understand physical structure. I am particularly interested in interpretable methods that link spectral variation to local atomic environments.

Selected Works:

• 📄 Multimodal Analysis of PDF and X-ray Absorption Spectra
  Na Narong, Zachko, Torrisi & Billinge, npj Computational Materials (Corresponding Author with Simon Billinge)
• 📄 Random Forest Models for Interpretable XANES Spectra
  Torrisi et al., npj Computational Materials

3. Electrochemical Materials and Device Informatics

I work on data-centric models for battery science — combining simulation, experiment, and statistical modeling to study device-level behavior and materials discovery for next-generation energy storage.

Selected Works:

• 📄 Interpretable Analysis of Battery Formation
  With the D3BATT Collaboration (Sun, Bazant, Chueh, et al.), Joule
• 📄 History-Agnostic Battery Degradation Inference
  Ansari, Torrisi, Trewartha, Sun, Journal of Energy Storage

My interests

I am passionate about science communication, funk music, and mentorship. While I was a Ph.D. student at Harvard, I was a resident tutor at Cabot House for four very happy years.

Older News

• May 24
  • The TM23 Dataset and Benchmarking paper is finally live- many thanks to my co-first author, Cameron Owen, and the rest of the Kozinsky group!
  • Excited to be hosting Killian Sheriff (a PhD student at MIT) as an intern at TRI this summer!
• April 24:
  • I gave two invited talks at MRS 2024 in Seattle, WA!
• March 24:
  • I gave an invited talk at Brookhaven National Laboratory in NY!
  • I gave a talk at the Columbia Data Science Institute
• December 23:
  • I presented at the Advanced Automotive Battery Conference in San Diego, CA!
  • My collaboration with Eli Gerber and Eun-ah Kim of Cornell University developing a tool called InterMatch is live now in Nature Communications!
  • Work by Bianca Baldasarri from the group of Chris Wolverton on a database and feature development for Oxygen Vacancy formation energy prediction is now live on Chemistry of Materials!
• September 23:
  • Look for a work by Mehrad Ansari, intern Summer ‘22, on machine learning for short-term battery agnostic degradation prediction.
• Aug 23:
  • August 23: I gave a lecture on the representation of materials at the SUNCAT summer school at Stanford University. It was an honor to present along other distinguished speakers and to have a chance to reach an audience of more early-career scientists!
• June 23:
  • I gave a talk at Argonne National Labs in their Machine Learning & Science seminar series!
  • I spoke at the Telluride Summer Science Conference in Colorado at their Machine Learning in Chemistry & Materials Science conference.
• May 23:
  • I will be hosting Viktoriia Baibakova (UC Berkeley) and Felix Jimenez (Texas A&M) as an intern mentor this summer at TRI!
  • A perspective paper I wrote with Shijing Sun (of TRI) and with many members of our consortium is now live on Applied Physics Letters- Machine Learning!
  • This website is started. Thanks for visiting!