Pascal Jutras-Dubé

PhD student (Computer Science), Purdue University — advised by Ruqi Zhang. I work on sampling and generative modeling through learned stochastic processes.

Currently into

• Diffusion LLM foundations
• Parallel decoding & efficient sampling
• Test-time alignment / controllability

Now seeking: Fall 2026 research internship.

Publications

P. Jutras-Dubé, J. Zhang, Z. Wang, R. Zhang. (2026) One-step diffusion samplers via self-distillation and deterministic flow. Annual conference on artificial intelligence and statisticss (AISTATS).

P. Punyamoorty, P. Jutras-Dubé, R. Zhang, V. Aggarwal, D. Conover, A. Bera. (2025) Dynamic obstacle avoidance through uncertainty-based adaptive planning with diffusion. International conference on intelligent robots and systems (IROS) [Co-first authors].

P. Jutras-Dubé, P. Pynadath, R. Zhang. (2025) Single-step consistent diffusion samplers. Frontiers in probabilistic inference workshop at ICLR.

P. Mesana, P. Jutras-Dubé, J. Crowe, G. Vial, G. Caporossi, S. Gambs. (2025) Measuring privacy/utility tradeoffs of format-preserving strategies for data release. Journal of Business Analytics.

P. Jutras-Dubé, M. B. Al-Khasawneh, Z. Yang, J. Bas, F. Bastin, C. Cirillo. (2024) Copula-based transferable models for synthetic population generation. Transportation Research Part C.

P. Jutras-Dubé, R. Zhang, A. Bera. (2024) Adaptive planning with generative models under uncertainty. International conference on intelligent robots and systems (IROS).

P. Mesana, P. Jutras-Dubé, J. Crowe, G. Vial, G. Caporossi. (2024) Evaluating the risk of re-identification in data release strategies: An attacker-centric approach. Hawaii international conference on system sciences (HICSS).