Sandbox environment for Windows, primarily designed for auto-judge systems for programmings competitions. Used from 2013 to 2024 in ANOI to let students who had access only to outdated dev environm…

A repo for transferable sampling of molecular systems

🏛️A research-friendly codebase for fast experimentation of single-agent reinforcement learning in JAX • End-to-End JAX RL

A jax-native environment interface

A simple, performant, and scalable Jax LLM!

Official implementation of DiscoGen, for "Procedural Generation of Algorithm Discovery Tasks in Machine Learning"

Accompanying code for "Discovering State-of-the-art Reinforcement Algorithms" Nature publication

A simple, performant and scalable JAX-based world modeling codebase.

Official implementation for "How Should We Meta-Learn Reinforcement Learning Algorithms?"

A JAX-Native Interface for Reinforcement Learning Environments

AIRA-dojo: a framework for developing and evaluating AI research agents

Parallel hyperparameter tuning with JAX

Linguini is a benchmark to measure a language model’s linguistic reasoning skills without relying on pre-existing language-specific knowledge, based on the International Linguistic Olympiad problems.

Library for efficient training and application of Machine Learning Interatomic Potentials (MLIP)

Official Implementation of "Can Learned Optimization Make Reinforcement Learning Less Difficult"

Unified Implementations of Offline Reinforcement Learning Algorithms

Reward Evolution with Large Language Models using Human Feedback

Latent Program Network (from the "Searching Latent Program Spaces" paper)

Demonstrations of Loss of Plasticity and Implementation of Continual Backpropagation

Reinforcement learning on general 2D physics environments in JAX. ICLR 2025 Oral.

Cellular Automata Accelerated in JAX (Oral at ICLR 2025)

The fundamental package for scientific computing with Python.

JAX reimplementation of the DeepMind paper "Genie: Generative Interactive Environments"

BenchMARL is a library for benchmarking Multi-Agent Reinforcement Learning (MARL). BenchMARL allows to quickly compare different MARL algorithms, tasks, and models while being systematically ground…

Accelerated minigrid environments with JAX

Deep memory and sequence models in JAX

VMAS is a vectorized differentiable simulator designed for efficient Multi-Agent Reinforcement Learning benchmarking. It is comprised of a vectorized 2D physics engine written in PyTorch and a set …

A tool for aggregating and plotting MARL experiment data.

♟️ Vectorized RL game environments in JAX