MoE-ACT: Scaling Multi-Task Bimanual Manipulation with Sparse Language-Conditioned Mixture-of-Experts Transformers

• URL: http://arxiv.org/abs/2603.15265v1
• Date: Mon, 16 Mar 2026 13:33:59 GMT
• Title: MoE-ACT: Scaling Multi-Task Bimanual Manipulation with Sparse Language-Conditioned Mixture-of-Experts Transformers
• Abstract summary: We propose a lightweight multi-task imitation learning framework for bimanual manipulation.<n>MoE-ACT integrates sparse Mixture-of-Experts (MoE) modules into the Transformer encoder of ACT.<n>We show that MoE-ACT substantially improves multi-task performance.
• Score: 3.890941830250993
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: The ability of robots to handle multiple tasks under a unified policy is critical for deploying embodied intelligence in real-world household and industrial applications. However, out-of-distribution variation across tasks often causes severe task interference and negative transfer when training general robotic policies. To address this challenge, we propose a lightweight multi-task imitation learning framework for bimanual manipulation, termed Mixture-of-Experts-Enhanced Action Chunking Transformer (MoE-ACT), which integrates sparse Mixture-of-Experts (MoE) modules into the Transformer encoder of ACT. The MoE layer decomposes a unified task policy into independently invoked expert components. Through adaptive activation, it naturally decouples multi-task action distributions in latent space. During decoding, Feature-wise Linear Modulation (FiLM) dynamically modulates action tokens to improve consistency between action generation and task instructions. In parallel, multi-scale cross-attention enables the policy to simultaneously focus on both low-level and high-level semantic features, providing rich visual information for robotic manipulation. We further incorporate textual information, transitioning the framework from a purely vision-based model to a vision-centric, language-conditioned action generation system. Experimental validation in both simulation and a real-world dual-arm setup shows that MoE-ACT substantially improves multi-task performance. Specifically, MoE-ACT outperforms vanilla ACT by an average of 33% in success rate. These results indicate that MoE-ACT provides stronger robustness and generalization in complex multi-task bimanual manipulation environments. Our open-source project page can be found at https://j3k7.github.io/MoE-ACT/.

Related papers

• Dynamics-Aware Meta-Imitation for Generalization to Unseen Robotic Manipulation [17.382238044141783]
  We propose a framework for learning skills from demonstrations for robots.<n>By integrating meta-learning to construct a shared skill space, DAMI equips agents for rapid adaptation to novel tasks.<n>Our approach outperforms state-of-the-art baselines regarding direct inference on seen tasks and adaptation to unseen tasks.
  arXiv  Detail & Related papers  (2026-07-17T11:50:03Z)
• Semantically Structured Mixture-of-Experts for Compositional Robotic Manipulation [21.358519910897652]
  Diffusion-based policies have established a new standard for precise robotic manipulation but face a critical scalability bottleneck.<n>We introduce Semantically Structured Mixture-of-Experts Diffusion Policy (SMoDP) for compositional robotic manipulation.<n>Our approach outperforms representative diffusion and MoE-based baselines on multi-task benchmarks with significantly improved parameter efficiency.
  arXiv  Detail & Related papers  (2026-05-22T10:38:59Z)
• ATG-MoE: Autoregressive trajectory generation with mixture-of-experts for assembly skill learning [6.440473674041692]
  ATG-MoE is an end-to-end autoregressive trajectory generation method with mixture of experts for assembly skill learning from demonstration.<n>It integrates multi-modal feature fusion for scene and task understanding, autoregressive sequence modeling for temporally coherent trajectory generation, and a mixture-of-experts architecture for unified multi-skill learning.<n>We train and evaluate the proposed method on eight representative assembly skills from a pressure-reducing valve assembly task.
  arXiv  Detail & Related papers  (2026-03-19T15:28:17Z)
• Scaling Tasks, Not Samples: Mastering Humanoid Control through Multi-Task Model-Based Reinforcement Learning [49.82882141491629]
  We argue that effective online learning should scale the emphnumber of tasks, rather than the number of samples per task.<n>This regime reveals a structural advantage of model-based reinforcement learning.<n>We instantiate this idea with textbfEfficientZero-Multitask (EZ-M), a sample-efficient multi-task algorithm for online learning.
  arXiv  Detail & Related papers  (2026-03-02T05:07:43Z)
• UV-M3TL: A Unified and Versatile Multimodal Multi-Task Learning Framework for Assistive Driving Perception [71.19234323863314]
  We propose a framework to simultaneously recognize driver behavior, driver emotion, vehicle behavior, and traffic context.<n>Our framework incorporates two core components: dual-branch spatial channel multimodal embedding and adaptive feature-decoupled multi-task loss.<n>We evaluate our method on the AIDE dataset, and the experimental results demonstrate that UV-M3TL achieves state-of-the-art performance across all four tasks.
  arXiv  Detail & Related papers  (2026-02-02T03:35:24Z)
• Training One Model to Master Cross-Level Agentic Actions via Reinforcement Learning [42.1534425503333]
  CrossAgent is a unified agentic model that masters heterogeneous action spaces and autonomously selects the most effective interface for each step of a trajectory.<n>Experiments on over 800 tasks in the open-world Minecraft environment demonstrate that CrossAgent achieves state-of-the-art performance.
  arXiv  Detail & Related papers  (2025-12-10T14:52:29Z)
• HiMoE-VLA: Hierarchical Mixture-of-Experts for Generalist Vision-Language-Action Policies [83.41714103649751]
  Development of embodied intelligence models depends on access to high-quality robot demonstration data.<n>We present HiMoE-VLA, a novel vision-language-action framework tailored to handle diverse robotic data with heterogeneity.<n>HiMoE-VLA demonstrates a consistent performance boost over existing VLA baselines, achieving higher accuracy and robust generalizations.
  arXiv  Detail & Related papers  (2025-12-05T13:21:05Z)
• Contextual Attention Modulation: Towards Efficient Multi-Task Adaptation in Large Language Models [23.932127056475526]
  Large Language Models (LLMs) possess remarkable capabilities but struggle with multi-task adaptation.<n>We propose Contextual Attention Modulation (CAM), a novel mechanism that dynamically modulates the representations of self-attention modules in LLMs.<n>Our approach significantly outperforms existing approaches, achieving an average performance improvement of 3.65%.
  arXiv  Detail & Related papers  (2025-10-20T16:19:27Z)
• Transforming Vision Transformer: Towards Efficient Multi-Task Asynchronous Learning [59.001091197106085]
  Multi-Task Learning (MTL) for Vision Transformer aims at enhancing the model capability by tackling multiple tasks simultaneously.<n>Most recent works have predominantly focused on designing Mixture-of-Experts (MoE) structures and in tegrating Low-Rank Adaptation (LoRA) to efficiently perform multi-task learning.<n>We propose a novel approach dubbed Efficient Multi-Task Learning (EMTAL) by transforming a pre-trained Vision Transformer into an efficient multi-task learner.
  arXiv  Detail & Related papers  (2025-01-12T17:41:23Z)
• Sparse Diffusion Policy: A Sparse, Reusable, and Flexible Policy for Robot Learning [61.294110816231886]
  We introduce a sparse, reusable, and flexible policy, Sparse Diffusion Policy (SDP) SDP selectively activates experts and skills, enabling efficient and task-specific learning without retraining the entire model. Demos and codes can be found in https://forrest-110.io/sparse_diffusion_policy/.
  arXiv  Detail & Related papers  (2024-07-01T17:59:56Z)
• CLAS: Coordinating Multi-Robot Manipulation with Central Latent Action Spaces [9.578169216444813]
  This paper proposes an approach to coordinating multi-robot manipulation through learned latent action spaces that are shared across different agents. We validate our method in simulated multi-robot manipulation tasks and demonstrate improvement over previous baselines in terms of sample efficiency and learning performance.
  arXiv  Detail & Related papers  (2022-11-28T23:20:47Z)
• UPDeT: Universal Multi-agent Reinforcement Learning via Policy Decoupling with Transformers [108.92194081987967]
  We make the first attempt to explore a universal multi-agent reinforcement learning pipeline, designing one single architecture to fit tasks. Unlike previous RNN-based models, we utilize a transformer-based model to generate a flexible policy. The proposed model, named as Universal Policy Decoupling Transformer (UPDeT), further relaxes the action restriction and makes the multi-agent task's decision process more explainable.
  arXiv  Detail & Related papers  (2021-01-20T07:24:24Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.