ACE-Brain-0: Spatial Intelligence as a Shared Scaffold for Universal Embodiments

2026-03-03

AI Summary

This work addresses the challenge of balancing generalization and domain specialization in universal embodied intelligence across heterogeneous platforms—such as autonomous vehicles, robots, and drones—which is often hindered by long-tailed data distributions, gradient interference, and catastrophic forgetting. The authors propose ACE-Brain-0, a framework that leverages spatial intelligence as a universal cognitive scaffold across diverse embodiment modalities. It introduces a Scaffold-Specialize-Reconcile (SSR) training paradigm: first establishing a shared spatial cognition foundation, then developing domain-specific expert models, and finally enabling data-free model fusion. Additionally, Group Relative Policy Optimization (GRPO) is incorporated to enhance holistic decision-making. Evaluated on 24 benchmarks spanning spatial and embodied intelligence, ACE-Brain-0 achieves state-of-the-art or superior performance, effectively harmonizing generalizability with task-specific expertise.

Technology Category

• Intelligent Robots: Embodied AI
• Humans and AI: Human-Aware Planning and Behavior Prediction
• Cognitive Modeling & Cognitive Systems: Agent Architectures

Application Category

• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs

Abstract

Universal embodied intelligence demands robust generalization across heterogeneous embodiments, such as autonomous driving, robotics, and unmanned aerial vehicles (UAVs). However, existing embodied brain in training a unified model over diverse embodiments frequently triggers long-tail data, gradient interference, and catastrophic forgetting, making it notoriously difficult to balance universal generalization with domain-specific proficiency. In this report, we introduce ACE-Brain-0, a generalist foundation brain that unifies spatial reasoning, autonomous driving, and embodied manipulation within a single multimodal large language model~(MLLM). Our key insight is that spatial intelligence serves as a universal scaffold across diverse physical embodiments: although vehicles, robots, and UAVs differ drastically in morphology, they share a common need for modeling 3D mental space, making spatial cognition a natural, domain-agnostic foundation for cross-embodiment transfer. Building on this insight, we propose the Scaffold-Specialize-Reconcile~(SSR) paradigm, which first establishes a shared spatial foundation, then cultivates domain-specialized experts, and finally harmonizes them through data-free model merging. Furthermore, we adopt Group Relative Policy Optimization~(GRPO) to strengthen the model's comprehensive capability. Extensive experiments demonstrate that ACE-Brain-0 achieves competitive and even state-of-the-art performance across 24 spatial and embodiment-related benchmarks.

Problem

• embodied intelligence
• universal generalization
• spatial intelligence
• catastrophic forgetting
• heterogeneous embodiments

Innovation

• spatial intelligence
• universal embodied intelligence
• Scaffold-Specialize-Reconcile
• multimodal large language model
• model merging