mlsanityData to 14 Oct 2026

ConceptACT: Episode-Level Concepts for Sample-Efficient Robotic Imitation Learning

arXiv 2601.17135 ↗First posted 23 Jan 2026cs.LG cs.ROcited by 0

On the map: Offline RL and imitation learning from data → Robot manipulation imitation learning

Abstract (arXiv)

Imitation learning enables robots to acquire complex manipulation skills from human demonstrations, but current methods rely solely on low-level sensorimotor data while ignoring the rich semantic knowledge humans naturally possess about tasks. We present ConceptACT, an extension of Action Chunking with Transformers that leverages episode-level semantic concept annotations during training to improve learning efficiency. Unlike language-conditioned approaches that require semantic input at deployment, ConceptACT uses human-provided concepts (object properties, spatial relationships, task constraints) exclusively during demonstration collection, adding minimal annotation burden. We integrate concepts using a modified transformer architecture in which the final encoder layer implements concept-aware cross-attention, supervised to align with human annotations. Through experiments on two robotic manipulation tasks with logical constraints, we demonstrate that ConceptACT converges faster and achieves superior sample efficiency compared to standard ACT. Crucially, we show that architectural integration through attention mechanisms significantly outperforms naive auxiliary prediction losses or language-conditioned models. These results demonstrate that properly integrated semantic supervision provides powerful inductive biases for more efficient robot learning.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Robot manipulation imitation learning, one of 13 directions in Offline RL and imitation learning from data. Builds on 30 of its 44 references that are in the map. First posted on arXiv 23 Jan 2026.

Builds on

30 of 44 references.

Closest ancestors · 2

Same direction, newest first

  1. 2023FurnitureBench: Reproducible Real-World Benchmark for Long-Horizon Complex Manipulation22 May
  2. 2023Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardware23 Apr

Other references · 28

  1. 2017Attention Is All You Needcited by 60,361
  2. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  3. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 30,186
  4. 2013Auto-Encoding Variational Bayescited by 4,053
  5. 2024OpenVLA: An Open-Source Vision-Language-Action Modelcited by 2,887
  6. 2023Diffusion Policy: Visuomotor Policy Learning via Action Diffusioncited by 2,838
  7. 2017FiLM: Visual Reasoning with a General Conditioning Layercited by 2,333
  8. 2022RT-1: Robotics Transformer for Real-World Control at Scalecited by 2,061
  9. 2017Interpretability Beyond Feature Attribution: Quantitative Testing with Concept Activation Vectors (TCAV)cited by 1,319
  10. 2020Concept Bottleneck Modelscited by 1,032
  11. 2021Deep Reinforcement Learning at the Edge of the Statistical Precipicecited by 807
  12. 2018This Looks Like That: Deep Learning for Interpretable Image Recognitioncited by 737
16 more
  1. 2021What Matters in Learning from Offline Human Demonstrations for Robot Manipulationcited by 725
  2. 2018Towards Robust Interpretability with Self-Explaining Neural Networkscited by 505
  3. 2020Concept Whitening for Interpretable Image Recognitioncited by 225
  4. 2023Foundation Models in Robotics: Applications, Challenges, and the Futurecited by 216
  5. 2023Language-Driven Representation Learning for Roboticscited by 165
  6. 2023A Survey of Imitation Learning: Algorithms, Recent Developments, and Challengescited by 148
  7. 2020Language-Conditioned Imitation Learning for Robot Manipulation Taskscited by 136
  8. 2020Widening the Pipeline in Human-Guided Reinforcement Learning with Explanation and Context-Aware Data Augmentationcited by 42
  9. 2025RoboEval: Where Robotic Manipulation Meets Structured and Scalable Evaluationcited by 32
  10. 2024Entity-Centric Reinforcement Learning for Object Manipulation from Pixelscited by 27
  11. 2020Structure Preserving Stain Normalization of Histopathology Images Using Self-Supervised Semantic Guidancecited by 22
  12. 2023State2Explanation: Concept-Based Explanations to Benefit Agent Learning and User Understandingcited by 17
  13. 2020Learning from Demonstration with Weakly Supervised Disentanglementcited by 8
  14. 2023When Multi-Task Learning Meets Partial Supervision: A Computer Vision Reviewcited by 6
  15. 2024LICORICE: Label-Efficient Concept-Based Interpretable Reinforcement Learningcited by 5
  16. 2021Accelerating the Learning of TAMER with Counterfactual Explanationscited by 3

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025Learning Semantic Atomic Skills for Multi-Task Robotic Manipulation
  2. 2025Knowledge-Driven Imitation Learning: Enabling Generalization Across Diverse Conditions
  3. 2026Human-to-Robot Interaction: Learning from Video Demonstration for Robot Imitation
  4. 2025LLM Trainer: Automated Robotic Data Generation via Demonstration Augmentation using LLMs
  5. 2025Learning a Thousand Tasks in a Day
  6. 2025SKIL: Semantic Keypoint Imitation Learning for Generalizable Data-efficient Manipulation
  7. 2026Zero-WAM: In-Context World-Action Modeling from Human Videos for Open-Ended Task Generalization
  8. 2020Scalable Multi-Task Imitation Learning with Autonomous Improvement