iiinova — where your digital self keeps evolving

Levi Lelis

@programsynthesis.bsky.social

353 followers 486 following 81 posts

Associate Professor - University of Alberta Canada CIFAR AI Chair with Amii Machine Learning and Program Synthesis he/him; ele/dele 🇨🇦 🇧🇷 www.cs.ualberta.ca/~santanad

Posts Replies Media

Levi Lelis @programsynthesis.bsky.social · 02/07/2025

Previous work has shown that programmatic policies—computer programs written in a domain-specific language—generalize to out-of-distribution problems more easily than neural policies. Is this really the case? 🧵

294

Levi Lelis @programsynthesis.bsky.social · 13/06/2025

In other cases, where clustering seems unable to find relevant structure, the subgoal-based policies do not seem to harm the search, as in Sokoban problems.

100

Levi Lelis @programsynthesis.bsky.social · 13/06/2025

The empirical results are strong when clustering effectively detects the problem's underlying structure.

100

Levi Lelis @programsynthesis.bsky.social · 13/06/2025

🧵1/ New paper! 📄 Subgoal-Guided Policy Heuristic Search with Learned Subgoals, led by my PhD student @tuero.ca. arxiv.org/pdf/2506.07255 This paper follows the Levin tree search (LTS) research line and focuses on learning subgoal-based policies.

153

Levi Lelis @programsynthesis.bsky.social · 23/05/2025

🧵1/ New paper! 📄 InnateCoder: Learning Programmatic Options with Foundation Models This is Rubens Moraes' final chapter of his PhD thesis from Universidade Federal de Viçosa, Brazil, in collaboration with Quazi Sadmine and Hendrik Baier. arXiv: arxiv.org/abs/2505.12508

163

Levi Lelis @programsynthesis.bsky.social · 21/05/2025

Are you interested in programmatic representations in machine learning? Please consider submitting your work to these workshops: one at ICML and another at RLC. 🧵

260

Levi Lelis @programsynthesis.bsky.social · 04/02/2025

I am re-reading "Learning Differentiable Programs with Admissible Neural Heuristics." This is still one of my favorite papers from the past 5–6 years. arxiv.org/pdf/2007.12101

120

Levi Lelis @programsynthesis.bsky.social · 26/12/2024

I still believe that visual perception could play a key role in solving the ARC-AGI challenge. The fact that VLMs currently perform worse than LLMs doesn’t necessarily prove the challenge isn’t visual. It could be that VLMs lack the bias needed to solve the visual ARC riddles. x.com/fchollet/sta...

020

Levi Lelis @programsynthesis.bsky.social · 09/12/2024

Just arrived in Vancouver for our #NeurIPS2024 tutorial on programmatic RL and code generation. Come join us tomorrow.

081