Kamala Varma

IN PROGRESS

• Varma, K. & Dumitraș, T. MET-FL: Multi-Exit Training for Efficient Heterogeneous Federated Learning. (submitted to ICML’25)

BOOK CHAPTERS

• Zhou, Y., Baracaldo, N., Anwar, A., & Varma, K. (2022, July). Dealing with Byzantine Threats to Neural Networks. Federated Learning: A Comprehensive Overview of Methods and Applications, SPRINGER NATURE, S.l., 2022, pp. 391-414.

CONFERENCE PROCEEDINGS PAPERS

• Upadhyay, N., Marupudi, V., Varma, K., & Varma, S. (2025, March). Alignment of CNN and Human Judgments of Geometric and Topological Concepts. Association for the Advancement of Artificial Intelligence (AAAI 2025).

• Varma, K., Numanoğlu, A., Kaya, Y., & Dumitraș, T. (2024, April). Understanding, Uncovering, and Mitigating the Causes of Inference Slowdown for Language Models. IEEE Conference on Secure and Trustworthy Machine Learning (SaTML 2024).

• Varma, K., Zhou, Y., Baracaldo, N., & Anwar, Ali. (2021, September). LEGATO: A LayerwisE Gradient AggregaTiOn Algorithm for Mitigating Byzantine Attacks in Federated Learning. IEEE Cloud Computing 2021.

• Fisher, D., Markert, E., Roberts, A., & Varma, K. (2019, June). Region radio: An AI that finds and tells stories about places. In K. Grace, M. Cook, D. Ventura, M& M. L. Maher (Eds.), Proceedings of the 10th International Conference on Computational Creativity (ICCC 2019), pp. 336-340. Charlotte, NC: Association for Computational Creativity.

• Varma, K., Guy, S. J., & Interrante, V. (2017, November). Assessing the relevance of eye gaze patterns during collision avoidance in virtual reality. In R. W. Lindeman, G. Bruder, & D. Iwai (Eds.), Proceedings of the International Conference on Artificial Reality & Eurographics Symposium on Virtual Environments (ICAT-EGVE-2017), pp. 149-152. Adelaide, Australia: Eurographics Association.

WORKSHOP PAPERS

• Varma, K., Diao, E., Roosta, T., Ding, J., & Zhang, T. (2023, June). Once-for-All Federated Learning: Learning From and Deploying to Heterogeneous Clients. In International Workshop on Federated Learning for Distributed Data Mining (FL4Data-Mining, 2023).

• Zhou, Y., Varma, K., & Baracaldo, N. Defending against adversarial attacks in federated learning. Published 1/11/2024.

• Anwar, A., Zawad, S., Zhou, Y., Baracaldo, N., Varma, K., Abay, A., Chuba, E., Ong, Y., & Ludwig, H. Tokenized Federated Learning. Published 1/19/2023.

• Zhou, Y., Baracaldo, N., Varma, K., Anwar, A., & Zawad, S. Byzantine-Robust Federated Learning. Published 12/24/2024.

Meta Machine Learning Engineer Intern — Summer 2024, Meta (Menlo Park, CA) (click to expand)

Manager: Gang Cheng
Worked on a label denoising project to improve Meta's TEXI model, which generates text embeddings optimized for integrity-related tasks (e.g., classifying text based on integrity violations). My project resulted in a ~6% performance improvement, and I wrote a guide detailing how the label denoising process can be applied to other embedding models across different modalities (e.g., image or video).

Research Mentor — Summer 2023, University of Maryland (College Park, MD) (click to expand)

Mentee: Arda Numanoğlu
Mentored an undergraduate student in the Summer at MC2 internship program. The intern co-authored a paper accepted to SaTML 2024.

Amazon Applied Scientist Intern — Summer 2022, Amazon (Boston, MA) (click to expand)

Team: Alexa Invocation
Manager: Dr. Jie Ding
Mentor: Dr. Tanya Roosta
Developed a method for federated learning systems involving clients with heterogeneous resource constraints. Authored a paper on this method, accepted to the International Workshop on Federated Learning for Distributed Data Mining.

IBM Graduate Research Intern — Summer 2021, IBM Almaden Research Center (San Jose, CA) (click to expand)

Team: AI Security and Privacy Solutions
Manager: Dr. Nathalie Baracaldo
Mentor: Dr. Yi Zhou
Analyzed Byzantine-robust aggregation algorithms in federated learning, developed a new defense, and co-applied for a patent. Studied certifiable robustness in federated learning contexts.

IBM Graduate Research Intern — Summer 2020, IBM Almaden Research Center (San Jose, CA) (click to expand)

Team: AI Security and Privacy Solutions
Manager: Dr. Nathalie Baracaldo
Mentor: Dr. Yi Zhou
Developed a Byzantine-robust aggregation algorithm for federated learning. Co-authored a paper accepted to IEEE Cloud 2021 and filed a patent (now granted).

IBM Undergraduate Research Intern — Summer 2019, IBM Almaden Research Center (San Jose, CA) (click to expand)

Team: AI Security and Privacy Solutions
Manager: Dr. Nathalie Baracaldo
Mentor: Dr. Yi Zhou
Researched clean-label poisoning attacks on deep neural networks and explored detection/mitigation methods.

NSF Computational Sustainability Network Intern — Summer 2018, Vanderbilt University (Nashville, TN) (click to expand)

Mentor: Dr. Douglas Fisher
Worked on an AI-driven storytelling tool focused on environmental topics using spatial context.

Big Data NSF REU Fellow — Summer 2017, University of Minnesota (Minneapolis, MN) (click to expand)

Mentor: Dr. Victoria Interrante
Conducted research on eye gaze patterns during collision avoidance in virtual reality environments.

REVIEWER EXPERIENCE

• 2025: International Conference on Machine Learning (ICML)

• 2025: International Conference on Learning Representations (ICLR)

• 2024: Conference on Neural Information Processing Systems (NeurIPS)

• 2023: Journal of Systems Architecture (JSA)

TEACHING ASSISTANTSHIPS

• Spring 2025: Object-Oriented Programming II (CMSC132)

• Fall 2024: Object-Oriented Programming I (CMSC131)

• Spring 2024: Algorithms (CMSC351)

• Fall 2022: Introduction to Data Science (CMSC320)

• Spring 2022: Discrete Structures (CMSC250)

• Fall 2021: Introduction to Data Science (CMSC320)

• Spring 2021: Algorithms (CMSC351)

• Fall 2020: Introduction to Artificial Intelligence (CMSC421)