GPU prices aggregator for cloud providers

Tokamax: A GPU and TPU kernel library.

Open-source framework for the research and development of foundation models.

Secondary Findings tool

Code for implementing central flows

Legible, Scalable, Reproducible Foundation Models with Named Tensors and Jax

a Jax quantization library

[NeurIPS 2025 Spotlight] Reasoning Environments for Reinforcement Learning with Verifiable Rewards

Official code release for "SuperBPE: Space Travel for Language Models"

Reference implementation and examples of the CuTe Layout representation and algebra.

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

Repository for measuring the volumes of minima in different datasets to test the volume and flat minima hypotheses.

📄 Forgetting in Language Models: Capacity, Optimization, and Self-Generated Replay

breseq is a computational pipeline for finding mutations relative to a reference sequence using high-throughput DNA resequencing data. It is intended for haploid microbial genomes (<20 Mb). breseq …

Second Order Optimization and Curvature Estimation with K-FAC in JAX.

A kernel library written in tilelang

Pytorch implementations of Elastic Weight Consolidation (EWC)

PyTorch-native post-training at scale

Reverb is an efficient and easy-to-use data storage and transport system designed for machine learning research

PyTorch implementation of various methods for continual learning (XdG, EWC, SI, LwF, FROMP, DGR, BI-R, ER, A-GEM, iCaRL, Generative Classifier) in three different scenarios.

Code accompanying the paper "Understanding Regularisation Methods for Continual Learning", https://arxiv.org/pdf/2006.06357.pdf

Code to accompany our paper "Continual Learning Through Synaptic Intelligence" ICML 2017

Local and global SPMD type system supporting erasure