https://mayankbakshi.github.io/Homepage/
The LIMITS Lab (Limits of Information and Trusted Distributed Systems) at Northern Arizona University (NAU), led by Prof. Mayank Bakshi, has an opening for a fully funded PhD student starting Fall 2026. The student will work on the theory and algorithms for distributed systems, with emphasis on efficiency and trustworthiness.
Potential projects may include, but are not limited to:
- Distributed and decentralized ML: communication-efficient training, federated/distributed optimization, and statistical limits
- Fine-tuning at scale: communication- and compute-aware fine-tuning of modern models, including LLMs/foundation models
- Adversarial robustness: Byzantine/poisoning-resilient learning and inference, security metrics and tradeoffs
- Error correction & coding for learning: reliability/efficiency mechanisms for distributed learning and inference pipelines
- Physical-layer security: adversarial channel/state models and information-theoretic security
- Structured recovery: sparsity/low-rank models for estimation and inference
Desired background: Applicants should have strong preparation in areas such as probability and linear algebra. Prior experience in networks, information theory, or distributed learning is helpful but not required. Students from EE, CS, Math, or related disciplines are encouraged to apply.
Mentorship & collaboration: Students will receive close mentorship and have opportunities for interdisciplinary collaboration across NAU (e.g., computing, engineering, mathematics) as well as with research groups at universities in the U.S. and internationally.
Location: NAU is located in Flagstaff, Arizona, a scenic mountain town with easy access to outdoor activities and the Grand Canyon region.
How to apply: Please apply through NAU’s graduate admissions process and email Prof. Bakshi with a brief introduction and your CV.
Lab homepage:
Contact: [email protected]
