Vignesh Kothapalli

vk_prof_pic_headshot.jpg

I am a second-year CS PhD student at Stanford University, advised by Prof. Jure Leskovec. I am interested in building foundation models over structured data (tables, databases, and graphs) and unstructured data (text). I see these models as the building blocks of continually learning decision-making systems. I co-lead the Stanford Tabular and Relational (STAR) Project with Rishabh Ranjan. My research is supported by the Stanford School of Engineering Fellowship.

In summer 2026, I interned at NVIDIA, where I worked on Kumo Tabular, a state-of-the-art open foundation model that makes predictions on new tables in a single forward pass, with no training or tuning. Previously I was a senior ML engineer at LinkedIn AI, building LLM-based foundation models for recommendation. I have also contributed to TensorFlow and maintained TensorFlow-IO at IBM.

I earned my MSc in Computer Science at NYU Courant, where I was advised by Prof. Joan Bruna in the Math and Data Group. I also hold a B.Tech in Electronics and Communication Engineering from IIT Guwahati.

Email / Scholar / GitHub / LinkedIn / X

Recent Updates

Show moreShow less

Publications

  1. RT-J: Large-Scale Pretraining of Relational Transformers for Context-Efficient Predictions
    Rishabh Ranjan, Vignesh Kothapalli, Harshvardhan Agarwal, Charilaos Kanatsoulis, Roshan Upendra, Tom Palczewski, Carlos Guestrin, and Jure Leskovec
    In The Fortieth Annual Conference on Neural Information Processing Systems , 2026
    PDF Code Model Data Website
  2. Memorilla: Latent Semantic Memory for LLMs
    Adibvafa Fallahpour*, Parsa Idehpour*, Vignesh Kothapalli*, Nikita Mounier, Shirley Wu, Shayan Pardis, and Jure Leskovec
    In COLM Workshop on Context Beyond the Window: Persistent Knowledge in Language Models , 2026
    * Equal contribution
    PDF Code Data Website
  3. PluRel: Synthetic Data unlocks Scaling Laws for Relational Foundation Models
    Vignesh Kothapalli, Rishabh Ranjan, Valter Hudovernik, Vijay Prakash Dwivedi, Johannes Hoffart, Carlos Guestrin, and Jure Leskovec
    International Conference on Machine Learning, 2026
    PDF Talk Blog Model Data Website
  4. Distilling the Essence: Efficient Reasoning Distillation via Sequence Truncation
    Wei-Rui Chen, Vignesh Kothapalli, Ata Fatahibaarzi, Hejian Sang, Shao Tang, Qingquan Song, Zhipeng Wang, and Muhammad Abdul-Mageed
    In Findings of the Association for Computational Linguistics , 2026
    PDF Code
  5. To Think or Not to Think: The Hidden Cost of Meta-Training with Excessive CoT Examples
    Vignesh Kothapalli, Ata Fatahibaarzi, Hamed Firooz, and Maziar Sanjabi
    In Proceedings of the Association for Computational Linguistics , 2026
    PDF Code Website
  6. Balancing Learning Rates Across Layers: Exact Two-Step Dynamics and Optimal Scaling in Linear Neural Networks
    Tianyu Pang, Vignesh Kothapalli, Shenyang Deng, Haohui Wang, Dawei Zhou, and Yaoqing Yang
    International Conference on Machine Learning, 2026
    PDF Code
  7. Scaling Down, Serving Fast: Compressing and Deploying Efficient LLMs for Recommendation Systems
    Kayhan Behdin, Ata Fatahibaarzi, Qingquan Song, Yun Dai, Aman Gupta, Zhipeng Wang, Hejian Sang, Shao Tang, Gregory Dexter, Sirou Zhu, Siyu Zhu, Tejas Dharamsi, Vignesh Kothapalli, Zhoutong Fu, Yihan Cao, Pin-Lun Hsu, Fedor Borisyuk, Natesh S. Pillai, Luke Simon, and Rahul Mazumder
    In Conference on Empirical Methods in Natural Language Processing (Industry Track) , 2025
    Oral PDF Code
  8. CoT-ICL Lab: A Synthetic Framework for Studying Chain-of-Thought Learning from In-Context Demonstrations
    Vignesh Kothapalli, Hamed Firooz, and Maziar Sanjabi
    In Proceedings of the Association for Computational Linguistics , 2025
    PDF Code Website
  9. From Spikes to Heavy Tails: Unveiling the Spectral Evolution of Neural Networks
    Vignesh Kothapalli, Tianyu Pang, Shenyang Deng, Zongmin Liu, and Yaoqing Yang
    Transactions on Machine Learning Research, 2025
    PDF
  10. Liger-Kernel: Efficient Triton Kernels for LLM Training
    Pin-Lun Hsu, Yun Dai, Vignesh Kothapalli, Qingquan Song, Shao Tang, Siyu Zhu, Steven Shimizu, Shivam Sahni, Haowen Ning, Yanning Chen, and Zhipeng Wang
    Championing Open-source DEvelopment in ML Workshop @ ICML25, 2025
    PDF Talk Blog Code
  11. Can Kernel Methods Explain How the Data Affects Neural Collapse?
    Vignesh Kothapalli, and Tom Tirer
    Transactions on Machine Learning Research, 2025
    PDF Code
  12. A Neural Collapse Perspective on Feature Evolution in Graph Neural Networks
    Vignesh Kothapalli, Tom Tirer, and Joan Bruna
    Advances in Neural Information Processing Systems, 2023
    PDF Code Slides
  13. Randomized Schur Complement Views for Graph Contrastive Learning
    Vignesh Kothapalli
    In International Conference on Machine Learning , 2023
    PDF Code
  14. Neural Collapse: A Review on Modelling Principles and Generalization
    Vignesh Kothapalli
    Transactions on Machine Learning Research, 2023
    PDF
  15. Abnormal Event Detection on BMTT-PETS 2017 Surveillance Challenge
    Vignesh Kothapalli, Gaurav Yadav, and Amit Sethi
    In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition Workshops , 2017
    PDF Code