Research Group
Daniel Wesego
PhD Student, University of Illinois Chicago
Robust ML & Deep Generative Models
Selected Publications
- D. Wesego, P. Rooshenas. TARO: Temporal Adversarial Rectification Optimization using Diffusion Models as Purifiers. arXiv preprint arXiv:2605.08440, 2026.
- D. Wesego, P. Rooshenas. Multimodal ELBO with Diffusion Decoders. CVPR Workshop on Any-to-Any Multi-modal Learning, 2026.
- D. Wesego, P. Rooshenas. Adversary Aware Optimization for Robust Defense. NeurIPS, 2025.
- D. Wesego, P. Rooshenas. Score-Based Multimodal Autoencoders. Transactions on Machine Learning Research, 2024.
Sumanta Bhattacharyya
PhD Student, University of Illinois Chicago
LLM Reasoning, Alignment & Controllable Generation
Selected Publications
- S. Bhattacharyya, P. Rooshenas. Steered Generation via Gradient-Based Optimization on Sparse Query Features. arXiv preprint arXiv:2605.23040, 2026.
- S. Bhattacharyya, P. Rooshenas. Audited Reasoning Refinement: Fine-Tuning Language Models via LLM-Guided Step-Wise Evaluation and Correction. arXiv preprint arXiv:2509.12476, 2025.
- S. Bhattacharyya, P. Rooshenas, S. Naskar, S. Sun, M. Iyyer, A. McCallum. Energy-Based Reranking: Improving Neural Machine Translation Using Energy-Based Models. ACL, 2021.
Guorui Sang
PhD Student, University of Illinois Chicago
Physics-Informed Deep Generative Models and Probabilistic Fields
Selected Publications
- G. Sang, P. Rooshenas. Hamiltonian-Guided Diffusion Fields for Variable-Length Rigid-Arm Trajectory Generation. ICLR Workshop on Real-World Constrained and Preference-Aligned Flow and Diffusion-Based Models, 2026.