I’m a PhD student at MIT co-advised by Armando Solar-Lezama in EECS and Josh Tenenbaum in BCS. My research combines methods from programming languages (PL) research with machine learning to tackle problems in artificial intelligence.
Research Interests
My research interests center on program synthesis, probabilistic programming, and artificial intelligence. I’m particularly interested in neurosymbolic methods that bridge the machine learning and programming languages communities. I believe symbolic methods can augment neural methods to facilitate low-data learning, generalization, transfer learning, interpretability, and other desiderata.
I’m particularly interested in abstraction learning, as in Ellis et al.’s DreamCoder. I led follow up work building a tool called Stitch (paper & code) published at POPL 2023 that achieves a 1,000-10,000x speedup in abstraction learning over DreamCoder. I’m interested in exploring new applications of abstraction learning, and I’m particularly interested in its application to world modelling through probablistic programs.
I previously published as Matthew Bowers.
Google Scholar / CV / Github / Bluesky / Twitter / Email (mlbowers@csail.mit.edu)
Publications
- Lazy Knowledge Compilation for Discrete PPLs (Languages For Inference Workshop at POPL 2025).
Maddy Bowers*, Alexander K. Lew*, Joshua B. Tenenbaum, Vikash Mansinghka, Armando Solar-Lezama. - Toward Probabilistic Coarse-to-Fine Program Synthesis (Languages for Inference Workshop at POPL 2024).
Maddy Bowers*, Alexander K. Lew*, Vikash Mansinghka, Joshua B. Tenenbaum, Armando Solar-Lezama. - LILO: Learning Interpretable Libraries by Compressing and Documenting Code (ICLR 2024).
Gabriel Grand, Lionel Wong, Maddy Bowers, Theo X. Olausson, Muxin Liu, Joshua B. Tenenbaum, Jacob Andreas. - Language Models Can Teach Themselves to Program Better (ICLR 2023).
Patrick Haluptzok, Maddy Bowers, Adam Tauman Kalai. - Top-Down Synthesis for Library Learning (POPL 2023; code).
Maddy Bowers, Theo X. Olausson, Catherine Wong, Gabriel Grand, Joshua B. Tenenbaum, Kevin Ellis, Armando Solar-Lezama. - Representing Partial Programs With Blended Abstract Semantics (ICLR 2021).
Maxwell Nye, Yewen Pu, Maddy Bowers, Jacob Andreas, Joshua B. Tenenbaum, Armando Solar-Lezama. - Universal Reshaping of Arrested Colloidal Gels via Active Doping (The Journal of Chemical Physics 2020).
Stewart Mallory, Maddy Bowers, Angelo Cacciuto. - Active Sculpting of Colloidal Crystals (The Journal of Chemical Physics 2019).
Shibananda Das, Maddy Bowers, Clara Bakker, Angelo Cacciuto.