LAMP Logo
LAMP Workshop
ML-assisted theory
TTIC Logo
Workshop on ML-assisted theory

Learning-driven Algorithms and Machine-aided Proofs (LAMP)

A workshop on machine learning for algorithm design, analysis, and proof generation — bringing together researchers in algorithms, learning theory, optimization, verification, and automated reasoning.

August 6-7, 2026 Research talks, open problems, panels, and lightning talks Audience: algorithms, ML theory, automated reasoning

About the Workshop

Recent advances in machine learning are beginning to reshape how algorithms are designed, analyzed, and even proved correct. Beyond using learning as a heuristic speed-up, a growing body of work explores ML for algorithm design and proofs: learning-augmented algorithms with performance guarantees, data-driven synthesis of algorithmic strategies, and models that generate or verify formal reasoning such as proofs, invariants, and certificates.

These developments raise foundational questions at the interface of theory and practice: When can learned components provably improve worst-case guarantees or be used to give beyond worst-case guarantees? How should we formalize the interaction between data-driven predictions and classical algorithmic analysis? And to what extent can machine learning assist in producing explanations or proofs that are both correct and interpretable?

This workshop aims to bring together researchers from algorithms, learning theory, and automated reasoning to develop principled frameworks, identify common abstractions, and chart future directions for using machine learning in advancing theoretical computer science.

Organizers

Dravyansh Sharma

Dravyansh Sharma

TTI-Chicago

Sandeep Silwal

Sandeep Silwal

University of Wisconsin–Madison

Ellen Vitercik

Ellen Vitercik

Stanford University

Speakers (tentative)

We have an exciting line-up of speakers on this timely research topic! The following researchers have confired interest (subject to the schedules working out).

Maria-Florina Balcan

Maria-Florina Balcan

Carnegie Mellon University

Vincent Cohen-Addad

Vincent Cohen-Addad

Google

Dylan Foster

Dylan Foster

Microsoft

Anupam Gupta

Anupam Gupta

New York University

Piotr Indyk

Piotr Indyk

MIT

Ravi Kumar

Ravi Kumar

Google

Debmalya Panigrahi

Debmalya Panigrahi

Duke University

Call for Spotlights, Posters, and Open Problems

We invite submissions for spotlight talks, posters, and open problems on topics related to the theme of the workshop, including but not limited to:

Submission details: Fill out this Google form for spotlight talks and posters. For open problems related to the topics in this workshop, send a two-page (exlcuding references) PDF document to dravy[PLUS]lamp26[AT]ttic[DOT]edu. Deadline: April 30, 2026.

Who Should Attend?

The workshop is intended for researchers and students interested in the interplay between machine learning and theoretical computer science, including algorithms, optimization, online learning, automated reasoning, formal methods, verification, and learning-augmented decision making.

Planned activities include invited talks, spotlight presentations, posters, open-problem discussions, and collaborative sessions centered on ML-assisted theory.

Contact

For questions about the workshop, submissions, or logistics, please write to:
dravy[PLUS]lamp26[AT]ttic[DOT]edu