Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Implementation of Gamma-World: Generative Multi-Agent World Modeling Beyond Two Players
| Date | Stars |
|---|---|
| 2026-07-31 | 651 |
| 2026-08-01 | 651 |
| 2026-08-06 | 651 |
Today
— stars today
This week
— stars this week
This month
— stars this month
Momentum
0.0
growth rate 0.00%/day
<div align="center">
<img src="assets/nvidia-logo.png" width="260" alt="NVIDIA">
<h1>
✨Gamma-World: Generative Multi-Agent World Modeling<br>
Beyond Two Players✨
</h1>
<p align="center">
<a href="https://liuff19.github.io/">Fangfu Liu</a><sup>1,2*</sup>
<a href="https://www.cs.toronto.edu/~hekai/">Kai He</a><sup>1,3,4*</sup>
<a href="https://www.cs.toronto.edu/~shenti11/">Tianchang Shen</a><sup>1</sup>
<a href="https://www.linkedin.com/in/tianshi-cao-a23270b1/">Tianshi Cao</a><sup>1</sup>
<a href="https://www.cs.utoronto.ca/~fidler/">Sanja Fidler</a><sup>1,3,4</sup>
<a href="https://duanyueqi.github.io/">Yueqi Duan</a><sup>2</sup>
<a href="https://www.cs.utoronto.ca/~jungao/">Jun Gao</a><sup>1</sup>
<br>
<a href="https://discover.research.utoronto.ca/32914-igor-gilitschenski">Igor Gilitschenski</a><sup>3,4†</sup>
<a href="https://www.cs.utoronto.ca/~zianwang/">Zian Wang</a><sup>1†</sup>
<a href="https://xuanchiren.com/">Xuanchi Ren</a><sup>1†</sup>
<br>
<br>
<sup>1</sup>NVIDIA
<sup>2</sup>Tsinghua University
<sup>3</sup>University of Toronto
<sup>4</sup>Vector Institute
</p>
<a href="https://research.nvidia.com/labs/sil/projects/gamma-world/"><img src="https://img.shields.io/badge/Project-Page-Green"></a>
<a href="assets/gamma-world.pdf"><img src="https://img.shields.io/badge/Paper-PDF-b31b1b.svg"></a>
<a href="https://arxiv.org/abs/2605.28816"><img src="https://img.shields.io/badge/arXiv-2605.28816-b31b1b.svg"></a>
<a><img src="https://img.shields.io/badge/License-Apache--2.0-blue"></a>
</div>
<strong>γ-World:</strong> We introduce a generative multi-agent world model that rolls out a single shared environment for multiple independently controllable agents. γ-World supports permutation-symmetric agent conditioning with <strong>Simplex Rotary Agent Encoding</strong>, efficient cross-agent communication with <strong>Sparse Hub Attention</strong>, real-time <strong>24 FPS</strong> streaming with a distilled block-causal student, and zero-shot generalization from two to four players.
## 📢 News
- 🚀[06/16/2026] We release the γ-World code and training pipeline.
- 🚀[05/28/2026] We release γ-World with the project page, paper, videos, qualitative results, and method overview.
## 🌟 Overview
γ-World interactively generates coherent future frames from multi-agent actions while preserving shared-world consistency, scaling from multiplayer virtual games to real-world multi-robot environments.
https://github.com/user-attachments/assets/11a81855-5b51-4117-bfcd-ef07246e0a4e
## 📦 User Guide
- [Setup](docs/setup.md) — environment and dependencies
- [Inference](docs/inference.md) — run the bidirectional, causal, and causal-few-step models on your own init frames and actions
- [Training](docs/training.md) — convert a released checkpoint to DCP and train the teacher / causal / DMD stages
## 📖 Abstract
World models for interactive video generation have largely focused on single-agent settings, where future observations are generated from a single control signal. However, many generated environments require multi-agent interaction: multiple players, robots, or embodied agents act simultaneously within a shared space. Scaling world models to such settings requires a principled multi-agent design: agents should remain independently controllable, permutation-symmetric, and support efficient inference while maintaining consistency across time and perspectives.
We present <strong>γ-World</strong>, a generative multi-agent world model for interactive simulation. γ-World introduces <strongExcerpt of 8,054 characters
Read on GitHub
6
Would you bet a product on this? Bounded 0–100 and slow moving.
matched fp:8ff2b1cd38bffb5a, topic:multi-agent, desc:multi-agent, desc:multi agent
matched fp:8ff2b1cd38bffb5a, topic:robotics