I am a fourth-year undergraduate student at Purdue University in Computer Science Honors, Statistics Honors, and Mathematics (expected 2027).
I build fundamental frameworks that view structured prediction and reasoning from a generative lens, spanning probabilistic modeling, variational inference, and flows across continuous, discrete, and hybrid settings.
I currently focus on jump diffusion processes for discrete and continuous generation across text, image, and puzzle-solving tasks working with Professor Ruqi Zhang. I worked on simulated and real-world environments for multi-agent robotics at the CoRAL lab under Professor Ahmed Qureshi.
I completed internships at Amazon AGI in 2025 and 2026 where I developed benchmarking tools for Amazon Nova Sonic’s speech and text models.
Research Interests
- Generative models: latent reasoning in discrete and continuous diffusions, unifying diffusion, flows, bridges, and interpolants, reward steering, search over generation.
- Probabilistic ML: simulation-free sampling, MCMC, Langevin Dynamics, SDEs, Poisson processes, Fokker-Planck and continuity PDEs.
- Applications to robotics: TAMP, goal-conditioned diffusion policies.
News
- Jul 2026: Preprint for Sticky Jump Diffusions released on arXiv.
- May 2026: Returned to Amazon AGI in Sunnyvale, CA.
- Jan 2026: Graphs-of-Constraints MPC has been accepted at ICRA 2026!
- May 2025: Interned at Amazon AGI in Sunnyvale, CA.
Selected Publications
- Sticky Jump Diffusions: A Unifying View of Masked, Continuous, and Hybrid Diffusion. P. Jutras-Dubé, P. Pynadath, J. Lu, Y. Gao, R. Zhang. Preprint, 2026. arXiv
- Graphs-of-Constraints Model Predictive Control for Reactive Multi-agent Task and Motion Planning. A. Manganaris, J. Lu, A. H. Qureshi, S. Jagannathan. ICRA, 2026. arXiv · Demo
See all publications on the Publications page.