Nofit Segal

I am a PhD student in the joint program of the Department of Materials Science and Engineering (DMSE) and the Center for Computational Science and Engineering (CSE) at MIT, advised by Prof. Rafael Gomez-Bombarelli.

I am passionate about developing practical machine-learning tools for materials inverse design. My current research focuses on predicting crystal structures from spectroscopic data and designing representation-learning models whose algorithmic structure supports stronger out-of-distribution generalization for materials.

I received my B.Sc. in Materials Science and Engineering from the Technion - Israel Institute of Technology, supported by the Technion Rothschild Excellence Fellowship.

Email  /  Google Scholar  /  LinkedIn  /  Github

• Two workshop papers accepted to the RPS Workshop at NeurIPS 2026, one as a Spotlight Talk. (2026)
• Our work on learning lattice parameters from powder XRD using invariants has been accepted to Acta Crystallographica Section A. (2026)
• Our work on probing the crystal structure–diffraction mapping through a gradient-descent lens has been accepted to Digital Discovery. (2026)
• Our work on probing the crystal structure–diffraction mapping through a gradient-descent lens has been accepted as an Oral Presentation to AI4Mat at NeurIPS 2025. (October 2025)
• Our work on using reinforcement learning with energy-based rewards for stable materials generation has been accepted to AI4Science at NeurIPS 2025. (October 2025)
• Our work on out-of-distribution property prediction for materials and molecules has been accepted to npj Computational Materials. (September 2025)

	MLIP Embeddings Predict Extraordinary Materials Luis Carretero López, Nofit Segal, Sathya Edamadaka, Rafael Gomez-Bombarelli Representations for the Physical Sciences (RPS), NeurIPS 2026 Spotlight Talk
	Self-Supervised Pretraining for Instrument-Robust Representations of Scattering Data Darin Merchant, Nofit Segal, Miguel Steiner, Rafael Gomez-Bombarelli Representations for the Physical Sciences (RPS), NeurIPS 2026
	Towards Generating Stable Materials via Large Language Models with Reinforcement Learning Finetuning Zhang-Wei Hong*, Nofit Segal*, Aviv Netanyahu, Hoje Chun, Rafael Gomez-Bombarelli, Pulkit Agrawal The Reach and Limits of AI for Scientific Discovery (AI4Science), NeurIPS 2025 paper

* Indicates equal contribution

Site based on this template.