OpenClaw-RL: Train any agent simply by talking

GRPO training code which scales to 32xH100s for long horizon terminal/coding tasks. Base agent is now the top Qwen3 agent on Stanford's TerminalBench leaderboard.

Bash is all you need - A nano claude code–like 「agent harness」, built from 0 to 1

My Solution to Assignment 1 for Stanford CS336 - Language Modeling From Scratch

My learning notes for ML SYS.

Jr. AI Scientist, a SOTA AI Scientist

Machine Learning Engineering Open Book

The best ChatGPT that $100 can buy.

Jax/Flax rewrite of Karpathy's nanoGPT

[ACM MM 2026]⚡ZEUS accelerates your diffuser. Any modality. Any model. Any scheduler. https://yixiao-wang-stats.github.io/zeus/

[ICML 2025] Official Repo for Stability-guided Adaptive Diffusion Acceleration. 🚀🌙Accelerating off-the-shelf diffusion model with a unified stability criterion.

COLM2026

Benchmark for automated failure attributions in agentic systems (🏆 ICML 2025 Spotlight)

HippoMM: Hippocampal-inspired Multimodal Memory

DiCoDe: Diffusion-Compressed Deep Tokens for Autoregressive Video Generation with Language Models

A curated list of papers and resources on byte-based large language models (LLMs) — models that operate directly on raw bytes.

Beyond Language Models: Byte Models are Digital World Simulators

Implementation of MEGABYTE, Predicting Million-byte Sequences with Multiscale Transformers, in Pytorch

This is the official Python version of CoreInfer: Accelerating Large Language Model Inference with Semantics-Inspired Adaptive Sparse Activation.

Modern CUDA Learn Notes with PyTorch for Beginners, 200+ CUDA Kernels, Tensor Cores, HGEMM, FA-2 MMA.

📚A curated list of Awesome LLM/VLM Inference Papers with Codes: Flash-Attention, Paged-Attention, WINT8/4, Parallelism, etc.🎉

📖 This is a repository for organizing papers, codes and other resources related to unified multimodal models.

Machine Learning and Computer Vision Engineer - Technical Interview Questions