Skip to content
Preprint

ME-Brain-1.0: Memory, Cognition and Action for Evolving Embodied Intelligence

Sep 2026 · 2 citations · 48 references
Computer Science

TL;DR

MachEmbodied-Brain (ME-Brain) is introduced, a self-evolving embodied system organized around a closed loop of action execution, experience acquisition, experience evolution, and improved execution that shifts embodied intelligence from train-and-freeze to deploy-and-evolve without model retraining.

Abstract

Current embodied systems largely rely on pretrained capabilities that remain fixed after deployment, limiting their ability to learn from physical interaction. We introduce MachEmbodied-Brain (ME-Brain), a self-evolving embodied system organized around a closed loop of action execution, experience acquisition, experience evolution, and improved execution. Evolvable Memory consolidates multimodal trajectories into hierarchical, reusable experience; Cognitive Core transforms physical experience into transferable skills; and the Action Model combines event-driven keyframes, EventCell local-world prediction, and action-conditioned memory modulation to focus computation on decision-critical moments, regions, and historical evidence. Together, these modules shift embodied intelligence from train-and-freeze to deploy-and-evolve without model retraining. Cognitive Core outperforms the strongest comparison models by 8.2 and 9.6 points on embodied and agent benchmarks. The Action Model achieves 47.88% mean success on RoboMME, a 3.26-point improvement over the strongest baseline. On RoboDojo, it reaches a 21.51 mean Score and 16.03% success rate, exceeding $\pi_{0.5}$ by 10.10 and 9.12 points. On the six-task ME-RealBench, ME-Brain achieves a 69.5 mean Score and 66.7% success rate, outperforming DM0.5 by 12.8 and 11.7 points, respectively.

View source

Similar papers

Preprint Sep 2026

ME-VLM: A Unified VLM for Embodied Cognition and Agent Coordination

Physical AI requires models to ground visual and linguistic understanding in real-world environments while accounting for environmental constraints and execution feedback. We introduce MachEmbodied-VLM (ME-VLM), a unified vision-language model with two variants, 4B and 35B-A3B, that brings together embodied cognition a...

Foundation Model, L. Inc · 0 citations
Preprint Sep 2026

EmbodiedMemory-Bench: Benchmarking Embodied Memory for Long-Horizon Embodied Tasks

Under matched backbones, EMem achieves the best overall performance among the evaluated memory systems and improves both open-source and proprietary models, while EMem-8B further improves over its backbone.

Li-Zhou Liang, Xin-Yue Zhong, Miao Pan et al. · 1 citation
Open access Sep 2026

Human-like memory empowers embodied robots for long-term object navigation

HM-Nav is presented, a brain-inspired architecture that bridges this cognitive gap by integrating transient sensory inputs with an evolving internal model to enable long-term object navigation and suggests that emulating human-like memory structures is essential for achieving resilient, long-term autonomy in complex, o...

Ying Zhang, Ren-Jie Song, Hong-Liang Ren et al. · 1 citation
Preprint Aug 2026

Skills in Weights, Memory in Code: Hybrid Learning for Memory-Dependent Robot Manipulation

HyMeS, a hybrid learning framework that leverages the reasoning and memory-management capabilities of coding agents to steer a Markovian VLA for memory-dependent manipulation, is proposed, enabling data-efficient compositional generalization.

Yun-Hao Zhao, Zhen-Yang Ni, Haoyang Chen et al. · 0 citations
#artificial intelligence Review Sep 2026

World Models for Embodied Intelligence: From Plausible to Controllable to Actionable

This framework surveys manipulation, navigation, locomotion, autonomous driving, and general embodied learning, tracing technical progressions, clarifying capability requirements, and examining datasets, benchmarks, and evaluation protocols.

Nan-Jie Yao, Hao Wang, Chong Cheng et al. · 1 citation
Preprint Sep 2026

Memory as Plans: World-Action Modeling with Memory-Grounded Planning

MaP-WAM is introduced, a Memory-as-Plans framework that decomposes memory-dependent world-action modeling into memory-grounded planning and plan-conditioned execution, and uses long-term multimodal episodic context as planning-time evidence rather than repeatedly conditioning the executor on the full history.

Si-Zhe Zhao, Hao-Zhe Xie, Wei-Yu Zhao et al. · 3 citations

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.