Anvith Thudi

PhD student

Email

Papers

1. Fast Exact Unlearning of Fine-Tuning Data for LLMs
   Andrei Ioan Muresanu, Anvith Thudi, Michael R. Zhang, Nicolas Papernot
   In Proceedings of the 42nd International Conference on Machine Learning, Vancouver, Canada

   BibTeX

   @inproceedings{andrei2025fastconference,
     author = {Muresanu, Andrei Ioan and Thudi, Anvith and Zhang, Michael R. and Papernot, Nicolas},
     booktitle = {Proceedings of the 42nd International Conference on Machine Learning, Vancouver, Canada},
     title = {Fast Exact Unlearning of Fine-Tuning Data for LLMs},
     year = {2025}
   }
   

2. Leveraging Per-Instance Privacy for Machine Unlearning
   Nazanin Mohammadi Sepahvand, Anvith Thudi, Berivan Isik, Ashmita Bhattacharyya, Nicolas Papernot, Eleni Triantafillou, Daniel M. Roy, Gintare Karolina Dziugaite
   In Proceedings of the 42nd International Conference on Machine Learning, Vancouver, Canada

   BibTeX

   @inproceedings{nazanin2025leveragingconference,
     author = {Sepahvand, Nazanin Mohammadi and Thudi, Anvith and Isik, Berivan and Bhattacharyya, Ashmita and Papernot, Nicolas and Triantafillou, Eleni and Roy, Daniel M. and Dziugaite, Gintare Karolina},
     booktitle = {Proceedings of the 42nd International Conference on Machine Learning, Vancouver, Canada},
     title = {Leveraging Per-Instance Privacy for Machine Unlearning},
     year = {2025}
   }
   

3. Leveraging Per-Instance Privacy for Machine Unlearning
   Nazanin Mohammadi Sepahvand, Anvith Thudi, Berivan Isik, Ashmita Bhattacharyya, Nicolas Papernot, Eleni Triantafillou, Daniel M. Roy, Gintare Karolina Dziugaite
   

   BibTeX

   @article{nazanin2025leveragingworkshop,
     author = {Sepahvand, Nazanin Mohammadi and Thudi, Anvith and Isik, Berivan and Bhattacharyya, Ashmita and Papernot, Nicolas and Triantafillou, Eleni and Roy, Daniel M. and Dziugaite, Gintare Karolina},
     title = {Leveraging Per-Instance Privacy for Machine Unlearning},
     year = {2025}
   }
   

4. Selective Classification Via Neural Network Training Dynamics
   Stephan Rabanser, Anvith Thudi, Kimia Hamidieh, Adam Dziedzic, Israfil Bahceci, Akram Bin Sediq, Hamza Sokun, Nicolas Papernot
   

   Paper BibTeX

   @article{stephan2025selectivejournal,
     author = {Rabanser, Stephan and Thudi, Anvith and Hamidieh, Kimia and Dziedzic, Adam and Bahceci, Israfil and Sediq, Akram Bin and Sokun, Hamza and Papernot, Nicolas},
     title = {Selective Classification Via Neural Network Training Dynamics},
     year = {2025}
   }
   

5. Unlearnable Algorithms for In-context Learning
   Andrei Ioan Muresanu, Anvith Thudi, Michael R. Zhang, Nicolas Papernot
   

   Paper BibTeX

   @article{andrei2024unlearnablepreprint,
     author = {Muresanu, Andrei Ioan and Thudi, Anvith and Zhang, Michael R. and Papernot, Nicolas},
     title = {Unlearnable Algorithms for In-context Learning},
     year = {2024}
   }
   

6. From Differential Privacy to Bounds on Membership Inference: Less can be More
   Anvith Thudi, Ilia Shumailov, Franziska Boenisch, Nicolas Papernot
   

   Paper BibTeX

   @article{anvith2024fromjournal,
     author = {Thudi, Anvith and Shumailov, Ilia and Boenisch, Franziska and Papernot, Nicolas},
     title = {From Differential Privacy to Bounds on Membership Inference: Less can be More},
     year = {2024}
   }
   

7. Gradients Look Alike: Sensitivity is Often Overestimated in DP-SGD
   Anvith Thudi, Hengrui Jia, Casey Meehan, Ilia Shumailov, Nicolas Papernot
   In Proceedings of the 33rd USENIX Security Symposium

   Paper BibTeX

   @inproceedings{anvith2024gradientsconference,
     author = {Thudi, Anvith and Jia, Hengrui and Meehan, Casey and Shumailov, Ilia and Papernot, Nicolas},
     booktitle = {Proceedings of the 33rd USENIX Security Symposium},
     title = {Gradients Look Alike: Sensitivity is Often Overestimated in DP-SGD},
     year = {2024}
   }
   

8. Proof-of-Learning is Currently More Broken Than You Think
   Congyu Fang, Hengrui Jia, Anvith Thudi, Mohammad Yaghini, Christopher A. Choquette-Choo, Natalie Dullerud, Varun Chandrasekaran, Nicolas Papernot
   In Proceedings of the 8th IEEE European Symposium on Security and Privacy, Delft, Netherlands

   Paper BibTeX

   @inproceedings{congyu2023proofoflearningconference,
     author = {Fang, Congyu and Jia, Hengrui and Thudi, Anvith and Yaghini, Mohammad and Choquette-Choo, Christopher A. and Dullerud, Natalie and Chandrasekaran, Varun and Papernot, Nicolas},
     booktitle = {Proceedings of the 8th IEEE European Symposium on Security and Privacy, Delft, Netherlands},
     title = {Proof-of-Learning is Currently More Broken Than You Think},
     year = {2023}
   }
   

9. Training Private Models That Know What They Don’t Know
   Stephan Rabanser, Anvith Thudi, Abhradeep Thakurta, Krishnamurthy Dvijotham, Nicolas Papernot
   In Proceedings of the 37th Conference on Neural Information Processing Systems

   Paper BibTeX

   @inproceedings{stephan2023trainingconference,
     author = {Rabanser, Stephan and Thudi, Anvith and Thakurta, Abhradeep and Dvijotham, Krishnamurthy and Papernot, Nicolas},
     booktitle = {Proceedings of the 37th Conference on Neural Information Processing Systems},
     title = {Training Private Models That Know What They Don’t Know},
     year = {2023}
   }
   

10. On the Necessity of Auditable Algorithmic Definitions for Machine Unlearning
    Anvith Thudi, Hengrui Jia, Ilia Shumailov, Nicolas Papernot
    In Proceedings of the 31st USENIX Security Symposium

    Paper BibTeX

    @inproceedings{anvith2022onconference,
      author = {Thudi, Anvith and Jia, Hengrui and Shumailov, Ilia and Papernot, Nicolas},
      booktitle = {Proceedings of the 31st USENIX Security Symposium},
      title = {On the Necessity of Auditable Algorithmic Definitions for Machine Unlearning},
      year = {2022}
    }
    

11. Unrolling SGD: Understanding Factors Influencing Machine Unlearning
    Anvith Thudi, Gabriel Deza, Varun Chandrasekaran, Nicolas Papernot
    In Proceedings of the 7th IEEE European Symposium on Security and Privacy, Genoa, Italy

    Paper BibTeX

    @inproceedings{anvith2022unrollingconference,
      author = {Thudi, Anvith and Deza, Gabriel and Chandrasekaran, Varun and Papernot, Nicolas},
      booktitle = {Proceedings of the 7th IEEE European Symposium on Security and Privacy, Genoa, Italy},
      title = {Unrolling SGD: Understanding Factors Influencing Machine Unlearning},
      year = {2022}
    }
    

12. Proof-of-Learning: Definitions and Practice
    Hengrui Jia, Mohammad Yaghini, Christopher A. Choquette-Choo, Natalie Dullerud, Anvith Thudi, Varun Chandrasekaran, Nicolas Papernot
    In Proceedings of the 42nd IEEE Symposium on Security and Privacy, San Francisco, CA

    Paper BibTeX

    @inproceedings{hengrui2021proofoflearningconference,
      author = {Jia, Hengrui and Yaghini, Mohammad and Choquette-Choo, Christopher A. and Dullerud, Natalie and Thudi, Anvith and Chandrasekaran, Varun and Papernot, Nicolas},
      booktitle = {Proceedings of the 42nd IEEE Symposium on Security and Privacy, San Francisco, CA},
      title = {Proof-of-Learning: Definitions and Practice},
      year = {2021}
    }
    

13. SoK: Machine Learning Governance
    Varun Chandrasekaran, Hengrui Jia, Anvith Thudi, Adelin Travers, Mohammad Yaghini, Nicolas Papernot
    

    Paper BibTeX

    @article{varun2021sokpreprint,
      author = {Chandrasekaran, Varun and Jia, Hengrui and Thudi, Anvith and Travers, Adelin and Yaghini, Mohammad and Papernot, Nicolas},
      title = {SoK: Machine Learning Governance},
      year = {2021}
    }