As of 2018, I'm working in Sanjit Seshia's research group. My work focuses on algorithmic improvisation, as well as the theory and applications of model counting and uniform generation.

My CV is here; my papers are also organized by topic below. My ORCID identifier is ORCID iD iconorcid.org/0000-0002-9992-9965.

Recent Work

Reactive Control Improvisation. [full version] [experiments]
Daniel J. Fremont and Sanjit A. Seshia.
At CAV 2018 (the 30th International Conference on Computer-Aided Verification).

Scenic: Language-Based Scene Generation. [full version] [tech report]
Daniel J. Fremont, Xiangyu Yue, Tommaso Dreossi, Shromona Ghosh, Alberto L. Sangiovanni-Vincentelli, and Sanjit A. Seshia.
UC Berkeley EECS Tech Report, 2018. (plus extended arXiv preprint)

Control Improvisation. [arXiv version]
Daniel J. Fremont, Alexandre Donzé, and Sanjit A. Seshia.
arXiv preprint, 2017. (extends the FSTTCS 2015 paper)

Maximum Model Counting. [full version] [tool and benchmarks]
Daniel J. Fremont, Markus N. Rabe, and Sanjit A. Seshia.
At AAAI 2017 (the 31st AAAI Conference on Artificial Intelligence).

Specification Mining for Machine Improvisation with Formal Specifications. [doi]
Rafael Valle, Alexandre Donzé, Daniel J. Fremont, Ilge Akkaya, Sanjit A. Seshia, Adrian Freed, and David Wessel.
In Computers in Entertainment, Vol. 14, No. 3, 2016.

Learning and Visualizing Music Specifications Using Pattern Graphs. [proceedings]
Rafael Valle, Daniel J. Fremont, Ilge Akkaya, Alexandre Donzé, Adrian Freed, and Sanjit A. Seshia.
At ISMIR 2016 (the 17th International Society for Music Information Retrieval Conference).
(N.B. See the journal version of this paper immediately above.)

On the Hardness of SAT with Community Structure. [arXiv version]
Nathan Mull, Daniel J. Fremont, and Sanjit A. Seshia.
At SAT 2016 (the 19th International Conference on Theory and Applications of Satisfiability Testing).

Control Improvisation with Probabilistic Temporal Specifications. [arXiv version]
Ilge Akkaya, Daniel J. Fremont, Rafael Valle, Alexandre Donzé, Edward A. Lee, and Sanjit A. Seshia.
At IoTDI 2016 (the 1st International Conference on Internet-of-Things Design and Implementation).
(best paper award)

Constrained Sampling and Counting: Universal Hashing Meets SAT Solving. [arXiv version]
Kuldeep S. Meel, Moshe Vardi, Supratik Chakraborty, Daniel J. Fremont, Sanjit A. Seshia, Dror Fried, Alexander Ivrii, and Sharad Malik.
At the AAAI-16 Workshop on Beyond NP, 2016.

Control Improvisation. [arXiv version]
Daniel J. Fremont, Alexandre Donzé, Sanjit A. Seshia, and David Wessel.
At FSTTCS 2015 (the 35th IARCS Annual Conference on Foundations of Software Technology and Theoretical Computer Science).
(N.B. This preliminary version of the Control Improvisation paper has been substantially extended: see the 2017 version above.)

On Parallel Scalable Uniform SAT Witness Generation. [full version] [tool] [benchmarks]
Supratik Chakraborty, Daniel J. Fremont, Kuldeep S. Meel, Sanjit A. Seshia, and Moshe Y. Vardi. (alphabetical order)
At TACAS 2015 (the 21st International Conference on Tools and Algorithms for the Construction and Analysis of Systems).

Distribution-Aware Sampling and Weighted Model Counting for SAT. [arXiv version]
Supratik Chakraborty, Daniel J. Fremont, Kuldeep S. Meel, Sanjit A. Seshia, and Moshe Y. Vardi. (alphabetical order)
At AAAI 2014 (the 28th AAAI Conference on Artificial Intelligence).

Speeding Up SMT-Based Quantitative Program Analysis. [arXiv version]
Daniel J. Fremont and Sanjit A. Seshia.
At SMT 2014 (the 12th International Workshop on Satisfiability Modulo Theories).

Older Work

In the summer of 2012 I did a UROP, advised by Henry Cohn. The original topic was the computability of Newton fractals, but my research took a tangent into reachability under affine functions on the integers and related problems. I talked about my work at the RP12 workshop - you can find my slides here. A writeup of my results is available on the arXiv.

In the spring of 2012 I did a UROP with Shankar Raman on the history of notions of continuity and the infinite in mathematics and philosophy. My research was mostly on Cavalieri's theory of indivisibles, Leibniz's various theories on the composition of the continuum, and the early development of set theory.

In the summer of 2011 I did a UROP on tetrahedral numbers, advised by Abhinav Kumar. Most of my time was spent learning about the circle method, which unfortunately seems unlikely to be of much use in proving Pollock's conjecture that every positive integer is the sum of at most 5 tetrahedral numbers. The current best result, by G. L. Watson in 1952 (Sums of eight values of a cubic polynomial, in J. London Math. Soc.) using an ad hoc method, is that 8 tetrahedral numbers suffice.