Scholar
Vineet Jain
Google Scholar ID: VsHC_00AAAAJ
McGill University, Mila
Citations & Impact
Citations
164
H-index
6
i10-index
4
Publications
10
Co-authors
0
Publications
Resume (English only)
Academic Achievements
- Publications: 'Recursive Self-Aggregation Unlocks Deep Thinking in Large Language Models', 'Diffusion Tree Sampling: Scalable inference-time alignment of diffusion models'.
Research Experience
- Worked at Valence Labs and Bosch Center for Artificial Intelligence. Research focuses on designing RL algorithms that elicit multimodal behavior and improve decision-making under uncertainty, developing diffusion and energy-based generative models for efficient sampling in high-dimensional spaces, and applying RL principles to improve inference-time alignment of large-scale generative models.
Education
- Ph.D. candidate at McGill University and Mila, supervised by Siamak Ravanbakhsh; M.S. in Machine Learning from Carnegie Mellon University; B.Tech. from the Indian Institute of Technology Kharagpur. During undergraduate studies, spent two summers at Mila, supervised by Yoshua Bengio, and at the Laboratory of Computational Neuroscience at EPFL, supervised by Wulfram Gerstner.
Background
- Research interests include reinforcement learning, probabilistic inference, and generative modeling. Long-term goal is to build autonomous agents that reason about their environment, adapt in real time, and generalize to unseen tasks.
Miscellany
- Can be contacted via Email, Google Scholar, LinkedIn, Twitter, or GitHub.
Co-authors