# DETACH: Cross-domain Learning for Long-Horizon Tasks via Mixture of Disentangled Experts

arXiv: https://arxiv.org/abs/2508.07842 · HTML: https://mlsanity.com/paper/2508.07842/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- Builds on 36 of its 48 references that are in the map.
- First posted on arXiv 11 Aug 2025.

## Builds on (most cited first, 30 of 36)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Proximal Policy Optimization Algorithms](https://mlsanity.com/paper/1707.06347/) (2017)
- [OpenVLA: An Open-Source Vision-Language-Action Model](https://mlsanity.com/paper/2406.09246/) (2024)
- [π0: A Vision-Language-Action Flow Model for General Robot Control](https://mlsanity.com/paper/2410.24164/) (2024)
- [MLP-Mixer: An all-MLP Architecture for Vision](https://mlsanity.com/paper/2105.01601/) (2021)
- [3D-FRONT: 3D Furnished Rooms with layOuts and semaNTics](https://mlsanity.com/paper/2011.09127/) (2020)
- [Pushing Mixture of Experts to the Limit: Extremely Parameter Efficient MoE for Instruction Tuning](https://mlsanity.com/paper/2309.05444/) (2023)
- [EmbodiedScan: A Holistic Multi-Modal 3D Perception Suite Towards Embodied AI](https://mlsanity.com/paper/2312.16170/) (2023)
- [Scaling Up Dynamic Human-Scene Interaction Modeling](https://mlsanity.com/paper/2403.08629/) (2024)
- [Towards Understanding Cross and Self-Attention in Stable Diffusion for Text-Guided Image Editing](https://mlsanity.com/paper/2403.03431/) (2024)
- [One Policy to Control Them All: Shared Modular Policies for Agent-Agnostic Control](https://mlsanity.com/paper/2007.04976/) (2020)
- [SimPoE: Simulated Character Control for 3D Human Pose Estimation](https://mlsanity.com/paper/2104.00683/) (2021)
- [HumanoidBench: Simulated Humanoid Benchmark for Whole-Body Locomotion and Manipulation](https://mlsanity.com/paper/2403.10506/) (2024)
- [Unified Human-Scene Interaction via Prompted Chain-of-Contacts](https://mlsanity.com/paper/2309.07918/) (2023)
- [Optimus-1: Hybrid Multimodal Memory Empowered Agents Excel in Long-Horizon Tasks](https://mlsanity.com/paper/2408.03615/) (2024)
- [Continual Learning and Catastrophic Forgetting](https://mlsanity.com/paper/2403.05175/) (2024)
- [InterDreamer: Zero-Shot Text to 3D Dynamic Human-Object Interaction](https://mlsanity.com/paper/2403.19652/) (2024)
- [TokenHSI: Unified Synthesis of Physical Human-Scene Interactions through Task Tokenization](https://mlsanity.com/paper/2503.19901/) (2025)
- [ICLR: In-Context Learning of Representations](https://mlsanity.com/paper/2501.00070/) (2024)
- [Steve-Eye: Equipping LLM-based Embodied Agents with Visual Perception in Open Worlds](https://mlsanity.com/paper/2310.13255/) (2023)
- [Diffusion Policies for Out-of-Distribution Generalization in Offline Reinforcement Learning](https://mlsanity.com/paper/2307.04726/) (2023)
- [Learning Generalizable Feature Fields for Mobile Manipulation](https://mlsanity.com/paper/2403.07563/) (2024)
- [Composite Motion Learning with Task Control](https://mlsanity.com/paper/2305.03286/) (2023)
- [Optimus-2: Multimodal Minecraft Agent with Goal-Observation-Action Conditioned Policy](https://mlsanity.com/paper/2502.19902/) (2025)
- [InteractAnything: Zero-shot Human Object Interaction Synthesis via LLM Feedback and Object Affordance Parsing](https://mlsanity.com/paper/2505.24315/) (2025)
- [Optimus-3: Dual-Router Aligned Mixture-of-Experts Agent with Dual-Granularity Reasoning-Aware Policy Optimization](https://mlsanity.com/paper/2506.10357/) (2025)
- [ContactDexNet: Multi-fingered Robotic Hand Grasping in Cluttered Environments through Hand-object Contact Semantic Mapping](https://mlsanity.com/paper/2404.08844/) (2024)
- [Contrastive Modules with Temporal Attention for Multi-Task Reinforcement Learning](https://mlsanity.com/paper/2311.01075/) (2023)
- [Controlling diverse robots by inferring Jacobian fields with deep networks](https://mlsanity.com/paper/2407.08722/) (2024)
- [Disentangled Unsupervised Skill Discovery for Efficient Hierarchical Reinforcement Learning](https://mlsanity.com/paper/2410.11251/) (2024)

## Built on by (most cited first, 5 of 5)

- [Embodied AI: From LLMs to World Models](https://mlsanity.com/paper/2509.20021/) (2025)
- [InterPrior: Scaling Generative Control for Physics-Based Human-Object Interactions](https://mlsanity.com/paper/2602.06035/) (2026)
- [U2UData+: A Scalable Swarm UAVs Autonomous Flight Dataset for Embodied Long-horizon Tasks](https://mlsanity.com/paper/2509.00055/) (2025)
- [MetaWorld: Skill Transfer and Composition in a Hierarchical World Model for Grounding High-Level Instructions](https://mlsanity.com/paper/2601.17507/) (2026)
- [ComplexMimic: Human-Scene Interaction Imitation in Complex 3D Environments](https://mlsanity.com/paper/2607.02034/) (2026)

## Abstract (arXiv)

Long-Horizon (LH) tasks in Human-Scene Interaction (HSI) are complex multi-step tasks that require continuous planning, sequential decision-making, and extended execution across domains to achieve the final goal. However, existing methods heavily rely on skill chaining by concatenating pre-trained subtasks, with environment observations and self-state tightly coupled, lacking the ability to generalize to new combinations of environments and skills, failing to complete various LH tasks across domains. To solve this problem, this paper presents DETACH, a cross-domain learning framework for LH tasks via biologically inspired dual-stream disentanglement. Inspired by the brain's "where-what" dual pathway mechanism, DETACH comprises two core modules: i) an environment learning module for spatial understanding, which captures object functions, spatial relationships, and scene semantics, achieving cross-domain transfer through complete environment-self disentanglement; ii) a skill learning module for task execution, which processes self-state information including joint degrees of freedom and motor patterns, enabling cross-skill transfer through independent motor pattern encoding. We conducted extensive experiments on various LH tasks in HSI scenes. Compared with existing methods, DETACH can achieve an average subtasks success rate improvement of 23% and average execution efficiency improvement of 29%.