Debopam Sanyal

I am a PhD student in Computer Science [SCS] at Georgia Institute of Technology, where I work in the Systems for Artificial Intelligence Lab (SAIL) under the supervision of Prof. Alexey Tumanov. My current research interests are in efficient and robust machine learning for flexible edge deployment.

I received my MS in Computer Science and my BS in Computer Engineering with a minor in Mathematics, both from University of Illinois Urbana-Champaign, where I was advised by Prof. Sibin Mohan.

During the summers of 2025 and 2026, I interned in the Advanced Technology Group at Dolby Laboratories, where I worked on efficient inference for large visual models and large language models.
Previously, I spent a summer as a research intern at Cisco Research and two summers as a research intern in the Scientific Data Division at Lawrence Berkeley National Laboratory, where I worked on the ScienceSearch project. During my undergraduate studies, I spent two years as a part-time researcher at National Center for Supercomputing Applications (NCSA).

I have received several awards, including being selected to attend the prestigious Heidelberg Laureate Forum (HLF) in 2022, as well as receiving the Sustainable Research Pathways Fellowship and the Fiddler Innovation Fellowship.

Email: debopam [dot] sanyal [at] gatech [dot] edu
Office: KACB 3110

Resume  /  Google Scholar  /  Github  /  Twitter  /  LinkedIn

profile photo

All Publications

* = equal contribution
  • SVD-Attention: Efficiently Analyzing Attention Blocks in Transformers.
    Debopam Sanyal, Hongjie Chen, Riya Kuriyan, Arth Tak, Divya Kothandaraman, Akshay Mehra, Joshua Kimball, Alexey Tumanov.
    Conference on Empirical Methods in Natural Language Processing (EMNLP), 2026.
    [Abstract][Paper]

  • A Benchmark for Analyzing Cloud Storage Layouts.
    Debopam Sanyal, Hongjie Chen, Alexey Tumanov, Joshua Kimball.
    ACM SIGOPS Operating Systems Review (OpSysRev), 2026.
    [Paper]

  • Robust Federated Learning Under Real-World Client Churn.
    Dhruv Garg, Neha Lakhani, Debopam Sanyal, Myungjin Lee, Alexey Tumanov, Ada Gavrilovska.
    arXiv preprint, 2026.
    [Paper]

  • DPAR: Dynamic Patchification for Efficient Autoregressive Visual Generation.
    Divyansh Srivastava, Akshay Mehra*, Pranav Maneriker*, Debopam Sanyal, Vishnu Raj, Vijay Kamarshi, Fan Du, Joshua Kimball.
    IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2026.
    [Paper][Poster]

  • LayoutBench: Performance Benchmarking of Cloud Storage Layouts for Multimedia Data.
    Debopam Sanyal, Hongjie Chen, Alexey Tumanov, Joshua Kimball.
    ACM Workshop on Machine Learning and Systems (EuroMLSys), 2026.
    [Paper][Slides]

  • KLAS: Using Similarity to Stitch Neural Networks for Improved Accuracy-Efficiency Tradeoffs.
    Debopam Sanyal, Anantharaman Iyer, Alind Khare, Trisha Jain, Akshay Jajoo, Myungjin Lee, Clayton Kerce, Alexey Tumanov.
    International Conference on Learning Representations (ICLR), 2026.
    [Paper][Slides][Poster]

  • VillainNet: Targeted Poisoning Attacks Against SuperNets Along the Accuracy-Latency Pareto Frontier.
    David Oygenblik, Abhinav Vemulapalli, Animesh Agrawal, Debopam Sanyal, Alexey Tumanov, Brendan Saltaformaggio.
    ACM Conference on Computer and Communications Security (CCS), 2025.
    [Paper]

  • Client Availability in Federated Learning: It Matters!
    Dhruv Garg*, Debopam Sanyal*, Myungjin Lee, Alexey Tumanov, Ada Gavrilovska.
    ACM Workshop on Machine Learning and Systems (EuroMLSys), 2025.
    [Paper][Slides]

  • Sci-Key: A Keyword Extraction Pipeline for Scientific Documents.
    Anna Giannakou, Oluwamayowa Amusat, Debopam Sanyal, Lavanya Ramakrishnan.
    Knowledge-Based Systems (KBS), 2024.
    [Paper]

  • Pareto-Secure Machine Learning (PSML): Fingerprinting and Securing Inference Serving Systems.
    Debopam Sanyal, Jui-Tse Hung, Manav Agrawal, Prahlad Jasti, Shahab Nikkhoo, Somesh Jha, Tianhao Wang, Sibin Mohan, Alexey Tumanov.
    arXiv preprint, 2023.
    [Paper]

  • Improving Schedule Indistinguishability in Real-Time Systems.
    Debopam Sanyal.
    Preprint, 2022.
    [Paper]

  • Attacking Schedule Indistinguishability in Real-Time Systems.
    Debopam Sanyal.
    Master's Thesis, Dept. of Computer Science, University of Illinois Urbana-Champaign, 2022.
    [Paper]

  • Indistinguishability Prevents Scheduler Side Channels in Real-Time Systems.
    Chien-Ying Chen, Debopam Sanyal, Sibin Mohan.
    ACM Conference on Computer and Communications Security (CCS), 2021.
    [Paper][Slides]

  • Feature Selection Metrics: Similarities, Differences, and Characteristics of the Selected Models.
    Debopam Sanyal, Nigel Bosch, Luc Paquette.
    International Conference on Educational Data Mining (EDM), 2020.
    [Paper][Slides]

  • Optimizing Networking Approaches using P4 Programming.
    Debopam Sanyal.
    Bachelor's Thesis, Dept. of Electrical & Computer Engineering, University of Illinois Urbana-Champaign, 2020.
    [Paper][Slides]