COMBAT: Conditional World Models for Behavioral Agent Training

TL;DR

COMBAT introduces a real-time diffusion-based world model trained on Tekken 3 that learns dynamic, reactive opponent behaviors implicitly from single-player data, enabling interactive agent simulation without explicit supervision.

Contribution

The paper presents COMBAT, a novel diffusion transformer model that models reactive agents in real-time using implicit learning from limited data, advancing interactive environment simulation.

Findings

Successfully simulates reactive opponent behavior in Tekken 3

Learns complex behaviors without explicit policy supervision

Establishes new benchmarks for emergent agent responsiveness

Abstract

Recent advances in video generation have spurred the development of world models capable of simulating 3D-consistent environments and interactions with static objects. However, a significant limitation remains in their ability to model dynamic, reactive agents that can intelligently influence and interact with the world. To address this gap, we introduce COMBAT, a real-time, action-controlled world model trained on the complex 1v1 fighting game Tekken 3. Our work demonstrates that diffusion models can successfully simulate a dynamic opponent that reacts to player actions, learning its behavior implicitly. Our approach utilizes a 1.2 billion parameter Diffusion Transformer, conditioned on latent representations from a deep compression autoencoder. We employ state-of-the-art techniques, including causal distillation and diffusion forcing, to achieve real-time inference. Crucially, we observe the emergence of sophisticated agent behavior by training the model solely on single-player inputs, without any explicit supervision for the opponent's policy. Unlike traditional imitation learning methods, which require complete action labels, COMBAT learns effectively from partially observed data to generate responsive behaviors for a controllable Player 1. We present an extensive study and introduce novel evaluation methods to benchmark this emergent agent behavior, establishing a strong foundation for training interactive agents within diffusion-based world models.