UniDisc: A discrete diffusion model for joint multimodal generation, enabling controllable and efficient text-image synthesis, editing, and inpainting.

Agent S: an open agentic framework that uses computers like a human

Code for: "Long-Context Autoregressive Video Modeling with Next-Frame Prediction"

R1-like Computer-use Agent

A python module to repair invalid JSON from LLMs

Explore the Multimodal “Aha Moment” on 2B Model

High-performance inference framework for large language models, focusing on efficiency, flexibility, and availability.

SANA: Efficient High-Resolution Image Synthesis with Linear Diffusion Transformer

Open-source Chrome extension for AI-powered web automation. Run multi-agent workflows using your own LLM API key. Alternative to OpenAI Operator.

Block Diffusion: Interpolating Between Autoregressive and Diffusion Language Models

Code accompanying the paper "Generalized Interpolating Discrete Diffusion"

L1: Controlling How Long A Reasoning Model Thinks With Reinforcement Learning

Make Mac apps accessible for AI agents

Domain-specific language designed to streamline the development of high-performance GPU/CPU/Accelerators kernels

🐳 Efficient Triton implementations for "Native Sparse Attention: Hardware-Aligned and Natively Trainable Sparse Attention"

A multi-player tournament benchmark that tests LLMs in social reasoning, strategy, and deception. Players engage in public and private conversations, form alliances, and vote to eliminate each other

Interesting CUA environment with conventional games

PyTorch implementation of FractalGen https://arxiv.org/abs/2502.17437

SlamKit is an open source tool kit for efficient training of SpeechLMs. It was used for "Slamming: Training a Speech Language Model on One GPU in a Day"

Solve Visual Understanding with Reinforced VLMs