Monty is a sensorimotor learning framework based on the thousand brains theory of the neocortex.

PyTorch Code for Energy-Based Transformers paper -- generalizable reasoning and scalable learning

AI agents running research on single-GPU nanochat training automatically

[RSS 2026] Interactive World Simulator for Robot Policy Training and Evaluation

100M tokens. Infinite compute. Lowest val loss wins.

A platform for reproducible world model research and evaluation

You can easily calculate FVD, PSNR, SSIM, LPIPS for evaluating the quality of generated or predicted videos.

Official PyTorch Implementation of "Diffusion Transformers with Representation Autoencoders"

An open source library designed to provide community examples of Joint Embedding Predictive Architectures (JEPAs). It contains code and examples for learning representations from images, video, and…

A Foundation Model for Generalist Gaming Agents

SR-DiT Speedrunning ImageNet Diffusion

Super-lean library for large-scale graph data

Midway Network: Learning Representations for Recognition and Motion from Latent Dynamics

LLM101n: Let's build a Storyteller

Run rapid transformer experiments. Heavily inspired by nanochat

🍓 Build and train energy-based and diffusion models in PyTorch ⚡.

The best ChatGPT that $100 can buy.

AI/ML Resources For Beginners

Flash Attention Triton kernel with support for second-order derivatives

An implementation of PSGD-QUAD optimizer for PyTorch

Reference PyTorch implementation and models for DINOv3

Physics of Language Models: Part 4.2, Canon Layers at Scale where Synthetic Pretraining Resonates in Reality

A framework for few-shot evaluation of language models.

Running Jax in PyTorch Lightning

Personal Website :)