About me

I’m a PhD candidate in Electrical and Computer Engineering at Georgia Tech, advised by Prof. Amirali Aghazadeh. I work on AI/ML with a focus on mechanistic interpretability, agentic AI, and generative and diffusion models — and I apply these methods to scientific problems.

My research spans interpretability of protein language models, cryo-EM image reconstruction, and the analysis of NASA astrobiology data for understanding the origin of life. My recent work has appeared at ICML 2026, CVPR 2026, and in PNAS Nexus.

A short overview of ASTRA, our multi-agent AI for reconstructing prebiotic reaction networks (paper · project):

Research interests

• Mechanistic interpretability — understanding the internal circuits of neural networks and protein language models (ProtoMech).
• Agentic AI for science — multi-agent systems that generate and test scientific hypotheses (ASTRA, AstroAgents).
• Generative & diffusion models — inverse problems and priors for scientific imaging (cryoSENSE).
• ML for astrobiology — discriminating biotic from abiotic organics in meteorite and terrestrial samples (LifeTracer).

Selected publications

• ICML 2026 — Protein Circuit Tracing via Cross-layer Transcoders
• CVPR 2026 — cryoSENSE: Compressive Sensing on the Protein Cryo-EM Image Manifold
• PNAS Nexus (2025) — Discriminating Abiotic and Biotic Organics in Meteorite and Terrestrial Samples Using Machine Learning on Mass Spectrometry Data
• ICML 2026 Mechanistic Interpretability Workshop — Circuit Tracing in Autoregressive Protein Language Models · ProtoMech: Protein Circuit Tracing via Cross-layer Transcoders
• ChemRxiv (2026) — ASTRA: Autonomous Multi-Agent Reconstruction of Prebiotic Reaction Networks

See all publications →

See my publications and projects, or download my CV.