Vishwak Srinivasan

About me: I am a PhD student in the EECS department at the Massachusetts Institute of Technology, advised by Ashia Wilson, and also closely work with Andre Wibisono at Yale University. Previously, I was a student in the Machine Learning Department at Carnegie Mellon University from where I earned a Master of Science, and completed my undergraduate studies at the Indian Institute of Technology Hyderabad where I majored in Computer Science and Engineering and graduated with honors.

My Research Interests: Currently, I'm interested in problems involving connections between optimisation and sampling, differential privacy and optimal transport.

Publications: A list of all papers that I've authored / co-authored can be found on my Google Scholar page. Recent papers / preprints:

• Fast sampling from constrained spaces using the Metropolis-adjusted Mirror Langevin algorithm (link)

  Vishwak Srinivasan, Andre Wibisono, Ashia Wilson

  Accepted at COLT 2024

• High-accuracy sampling from constrained spaces with the Metropolis-adjusted Preconditioned Langevin Algorithm (link)

  Vishwak Srinivasan, Andre Wibisono, Ashia Wilson

  Accepted at ALT 2025

• The Gaussian Mixing Mechanism: Renyi Differential Privacy via Gaussian Sketches (link)

  Omri Lev, Vishwak Srinivasan^*, Moshe Shenfeld^*, Katrina Ligett, Ayush Sekhari, Ashia Wilson

  Accepted at NeurIPS 2025

• Designing Algorithms for Entropic Optimal Transport from an Optimisation Perspective (link)

  Vishwak Srinivasan^*, Qijia Jiang^*

  Accepted at ALT 2026

• Understanding two-scale criteria for Poincaré and log-Sobolev inequalities in the Euclidean case through Φ-entropies (link)

  Vishwak Srinivasan

  Pre-print

• Near-Optimal Private Linear Regression via Iterative Hessian Mixing (link)

  Omri Lev, Moshe Shenfeld, Vishwak Srinivasan, Katrina Ligett, Ayush Sekhari, Ashia Wilson

  Pre-print, under submission

Teaching: I am currently a TA for 6.7830 (prev. 6.435) - Bayesian Modelling and Inference [Spring 2026]. Previously, I have been a TA for the following subjects:

• at MIT: 6.7830 (prev. 6.435) - Bayesian Modelling and Inference [Spring 2025], 6.7810 (prev. 6.438) - Algorithms for Inference [Fall 2023], and
• at Carnegie Mellon University: 10-725 - Convex Optimisation [Spring 2020, Fall 2020].

Other professional service:

• EECS Communication Lab Fellow (2024-).
• Reviewer for conferences: NeurIPS (2021, 2022 [Top Reviewer], 2023 [Top Reviewer], 2024, 2025), ICML (2022, 2023, 2024, 2025, 2026), AISTATS (2025), COLT (2025, 2026), AAAI (2026).
• Affiliation with student bodies: EECS Graduate Student Association (2022-2023), MIT Graduate Student Council as EECS Department Representative (2022-2024), Ashdown House Coffee Hour Committee (2021-2023).
• Served on the MIT EECS PhD Admissions Committee (2024), CMU MLD MS Admissions Committee (2020).

Contact information: My e-mail address is vishwaks at mit dot edu.

Random musings: I like doing recreational maths, and below is a (new) listing of random mathematical facts that I come across / derive.

• Bounds on a recurrence relation (link)
• Bounds on a nested sum (link)
• Positive semi-definiteness of a 2x2 matrix (link)
• Inverse CDFs of log-concave distributions are fascinating (link)
• Schur complements have to be complimented (link)
• Proxies for estimating functionals (link)