My research involves connections between computability, model theory, probability, and physics, including the computability and complexity theory of Bayesian inference, the model theory of probabilistic structures, and the statistics of large graphs.
I am currently a Research Scientist in the MIT Probabilistic Computing Project. Previously at MIT I was an Instructor in Pure Mathematics 2008–2010, a Postdoctoral Fellow in the Computer Science and Artificial Intelligence Laboratory 2011–2013, and a Postdoctoral Associate in the Department of Brain and Cognitive Sciences 2013–2015. I was a Junior Researcher in the Mathematics Department of the University of Hawaii at Manoa 2010–2011, a Lyric Labs Visiting Fellow at Analog Devices 2013–2014, a Research Scientist at Gamalon Labs 2013–2016, and Chief Scientist at Remine 2017–2018.
The Fast Loaded Dice Roller: A near-optimal exact sampler for discrete probability distributions, with Feras Saad, Martin Rinard, and Vikash Mansinghka, in Proceedings of the 23rd International Conference on Artificial Intelligence and Statistics (AISTATS 2020), Proceedings of Machine Learning Research (PMLR) 108, 1036–1046, 2020. arXiv:2003.03830.
Optimal approximate sampling from discrete probability distributions, with Feras Saad, Martin Rinard, and Vikash Mansinghka, Proceedings of the ACM on Programming Languages 4, POPL, 36:1–36:31, 2020. arXiv:2001.04555.
Computability of algebraic and definable closure, with Nate Ackerman and Rehana Patel, in Proceedings of the Symposium on Logical Foundations Of Computer Science (LFCS 2020), LNCS Vol. 11972, 1–11, 2020.
Algorithmic barriers to representing conditional independence, with Nate Ackerman, Jeremy Avigad, Daniel Roy, and Jason Rute, in Proceedings of the 34th Annual ACM/IEEE Symposium on Logic in Computer Science (LICS 2019), 2019.
Feedback computability on Cantor space, with Nate Ackerman and Robert Lubarsky, Selected Papers of Logic in Computer Science (LICS) 2015 and 2016, Logical Methods in Computer Science 15, no. 2, 7:1–7:18, 2019. arXiv:1708.01139.
A family of exact goodness-of-fit tests for high-dimensional discrete distributions, with Feras Saad, Nate Ackerman, and Vikash Mansinghka, in Proceedings of the 22nd International Conference on Artificial Intelligence and Statistics (AISTATS 2019), Proceedings of Machine Learning Research (PMLR) 89, 1640–1649, 2019. arXiv:1902.10142.
The Beta-Bernoulli process and algebraic effects, with Sam Staton, Dario Stein, Hongseok Yang, Nate Ackerman, and Daniel Roy, in Proceedings of the 45th International Colloquium on Automata, Languages, and Programming (ICALP 2018), 141:1-141:15, 2018. arXiv:1802.09598.
A classification of orbits admitting a unique invariant measure, with Nate Ackerman, Aleksandra Kwiatkowska, and Rehana Patel, Annals of Pure and Applied Logic, 168, no. 1, 19–36, 2017. arXiv:1412.2735.
Invariant measures concentrated on countable structures, with Nate Ackerman and Rehana Patel, Forum of Mathematics Sigma 4, e17, 59 pp., 2016. arXiv:1206.4011.
Invariant measures via inverse limits of finite structures, with Nate Ackerman, Jaroslav Neetřil, and Rehana Patel, European Journal of Combinatorics 52, 248–289, 2016. arXiv:1310.8147.
Feedback Turing computability, and Turing computability as feedback, with Nate Ackerman and Robert Lubarsky, in Proceedings of the 30th Annual ACM/IEEE Symposium on Logic in Computer Science (LICS 2015), 523–534, 2015.
Towards common-sense reasoning via conditional simulation: legacies of Turing in Artificial Intelligence, with Daniel Roy and Joshua Tenenbaum, in Turing's Legacy: Developments from Turing's Ideas in Logic, ed. Rod Downey, ASL Lecture Notes in Logic 42, Cambridge University Press, 2014. arXiv:1212.4799.
Randomness extraction and asymptotic Hamming distance, with Bjørn Kjos-Hanssen, Selected Papers of the Ninth International Conference on Computability and Complexity in Analysis (CCA 2012), Logical Methods in Computer Science 9, no. 3, 1–14, 2013. arXiv:1008.0821.
A notion of a computational step for Partial Combinatory Algebras, with Nate Ackerman, in Proceedings of the 10th Annual Conference on Theory and Applications of Models of Computation (TAMC 2013), LNCS Vol. 7876, 133–143, 2013.
Posterior distributions are computable from predictive distributions, with Daniel Roy, in Proceedings of the 13th International Conference on Artificial Intelligence and Statistics (AISTATS 2010), Journal of Machine Learning Research (JMLR) Workshop and Conference Proceedings 9, 233–240, 2010.
Computable exchangeable sequences have computable de Finetti measures, with Daniel Roy, in Mathematical Theory and Computational Practice, Proceedings of Computability in Europe (CiE 2009), LNCS Vol. 5635, 218–231, 2009.
Models with High Scott Rank, PhD thesis, Harvard University, 2008.
System and method for relativistic statistical securities trading, with Alexander Wissner-Gross, U.S. Patent 8,635,133 (2014).