Mauricio Barba da Costa

I'm a research engineer at a stealth startup, where we work on technology at the intersection of distributed systems, databases, and AI. Previously, I completed my B.S. and M.Eng. at MIT.

Conference & Journal

A Closer Look at Power Sampling

Mauricio Barba da Costa, Ryan Cotterell

Preprint

Faster LLM Inference via Sequential Monte Carlo

Yahya Emara*, Mauricio Barba da Costa*, Chi-Chih Chang, Cameron Freer, Tim Vieira, Ryan Cotterell, Mohamed S. Abdelfattah

Preprint — paper · semianalysis · blog post

People Use Fast, Flat Goal-Directed Simulation to Reason about Novel Problems

Katherine M. Collins*, Cedegao E. Zhang*, Lionel Wong*, Mauricio Barba da Costa*, Graham Todd*, Adrian Weller, Samuel J. Cheyette, Thomas L. Griffiths, Joshua B. Tenenbaum

Nature, 2026 — paper · nature podcast (starting at 10:59)

Evaluating Language Models' Evaluations of Games

Katherine M. Collins, Cedegao E. Zhang, Graham Todd, Lance Ying, Mauricio Barba da Costa, Ryan Liu, Prafull Sharma, Adrian Weller, Ionatan Kuperwajs, Lionel Wong, Joshua B. Tenenbaum, Thomas L. Griffiths

ICLR, 2026 — paper

Workshop

Sequential Monte Carlo Program Synthesis with Refinement Proposals

Maddy Bowers, Mauricio Barba da Costa, Xiaoyan Wang, Joshua B. Tenenbaum, Vikash K. Mansinghka, Armando Solar-Lezama, Alexander K. Lew

LAFI workshop at POPL, 2026 — link

Improving Autoformalization via Cycle Consistency and Incremental Type-Checking Using Language-Model Probabilistic Programs

Mauricio Barba da Costa*, Fabian Zaiser*, Katherine M. Collins, Romir Patel, Timothy J. O'Donnell, Alexander K. Lew, Joshua B. Tenenbaum, Vikash K. Mansinghka, Cameron E. Freer

NeurIPS MATH-AI Workshop, 2025 — link

Automatic Geometry Theorem Proving Using Polynomial Elaboration

Mauricio Barba da Costa, Fabian Zaiser, Cameron Freer, Josh Tenenbaum, Vikash Mansinghka

ITP 2025 Lean Workshop — slides