mlsanityData to 14 Oct 2026

Temporal Representation Alignment: Successor Features Enable Emergent Compositionality in Robot Instruction Following

arXiv 2502.05454 ↗First posted 8 Feb 2025cs.RO cs.LGcited by 14

Not on the map. Why?

Abstract (arXiv)

Effective task representations should facilitate compositionality, such that after learning a variety of basic tasks, an agent can perform compound tasks consisting of multiple steps simply by composing the representations of the constituent steps together. While this is conceptually simple and appealing, it is not clear how to automatically learn representations that enable this sort of compositionality. We show that learning to associate the representations of current and future states with a temporal alignment loss can improve compositional generalization, even in the absence of any explicit subtask planning or reinforcement learning. We evaluate our approach across diverse robotic manipulation tasks as well as in simulation, showing substantial improvements for tasks specified with either language or goal images.

Key facts. Cited by 14 papers in the map (data through 14 Oct 2026). Builds on 65 of its 87 references that are in the map. First posted on arXiv 8 Feb 2025.

Builds on

65 of 87 references.

References · 65

  1. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
  2. 2020Language Models are Few-Shot Learnerscited by 32,419
  3. 2022PaLM: Scaling Language Modeling with Pathwayscited by 5,326
  4. 2023RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Controlcited by 3,080
  5. 2024OpenVLA: An Open-Source Vision-Language-Action Modelcited by 2,887
  6. 2022Do As I Can, Not As I Say: Grounding Language in Robotic Affordancescited by 2,348
  7. 2021Decision Transformer: Reinforcement Learning via Sequence Modelingcited by 1,546
  8. 2024Octo: An Open-Source Generalist Robot Policycited by 1,484
  9. 2021Offline Reinforcement Learning with Implicit Q-Learningcited by 1,293
  10. 2018Learning Latent Dynamics for Planning from Pixelscited by 1,264
  11. 2021Grounded Language-Image Pre-trainingcited by 1,156
  12. 2017Hindsight Experience Replaycited by 1,136
38 more
  1. 2019Advantage-Weighted Regression: Simple and Scalable Off-Policy Reinforcement Learningcited by 782
  2. 2022R3M: A Universal Visual Representation for Robot Manipulationcited by 694
  3. 2021CLIPort: What and Where Pathways for Robotic Manipulationcited by 622
  4. 2022ProgPrompt: Generating Situated Robot Task Plans using Large Language Modelscited by 611
  5. 2022BC-Z: Zero-Shot Task Generalization with Robotic Imitation Learningcited by 523
  6. 2022MineDojo: Building Open-Ended Embodied Agents with Internet-Scale Knowledgecited by 478
  7. 2020Learning Invariant Representations for Reinforcement Learning without Reconstructioncited by 436
  8. 2022VIMA: General Robot Manipulation with Multimodal Promptscited by 407
  9. 2022VIP: Towards Universal Visual Reward and Representation via Value-Implicit Pre-Trainingcited by 395
  10. 2016Successor Features for Transfer in Reinforcement Learningcited by 340
  11. 2016Hierarchical Deep Reinforcement Learning: Integrating Temporal Abstraction and Intrinsic Motivationcited by 286
  12. 2022Contrastive Learning as Goal-Conditioned Reinforcement Learningcited by 253
  13. 2024OGBench: Benchmarking Offline Goal-Conditioned RLcited by 238
  14. 2022Interactive Language: Talking to Robots in Real Timecited by 218
  15. 2024RT-H: Action Hierarchies Using Languagecited by 210
  16. 2020Language Conditioned Imitation Learning over Unstructured Datacited by 209
  17. 2023LIV: Language-Image Representations and Rewards for Robotic Controlcited by 188
  18. 2023Language-Driven Representation Learning for Roboticscited by 165
  19. 2019Goal-conditioned Imitation Learningcited by 162
  20. 2021The Surprising Effectiveness of Representation Learning for Visual Imitationcited by 154
  21. 2019Planning with Goal-Conditioned Policiescited by 149
  22. 2019Unsupervised State Representation Learning in Ataricited by 144
  23. 2023HIQL: Offline Goal-Conditioned RL with Latent States as Actionscited by 141
  24. 2023Q-Transformer: Scalable Offline Reinforcement Learning via Autoregressive Q-Functionscited by 131
  25. 2023Optimal Goal-Reaching Reinforcement Learning via Quasimetric Learningcited by 121
  26. 2020Learning to Generalize Across Long-Horizon Tasks from Human Demonstrationscited by 115
  27. 2022Does CLIP Bind Concepts? Probing Compositionality in Large Image Modelscited by 102
  28. 2021RRL: Resnet as representation for Reinforcement Learningcited by 91
  29. 2021Online and Offline Reinforcement Learning by Planning with a Learned Modelcited by 90
  30. 2021Consequences of Misaligned AIcited by 80
  31. 2022Pre-Training for Robots: Offline RL Enables Learning New Tasks from a Handful of Trialscited by 78
  32. 2021Learning Successor States and Goal-Dependent Values: A Mathematical Viewpointcited by 70
  33. 2022Can Foundation Models Perform Zero-Shot Task Specification For Robot Manipulation?cited by 70
  34. 2020Learning to Recombine and Resample Data for Compositional Generalizationcited by 65
  35. 2023"No, to the Right" -- Online Language Corrections for Robotic Manipulation via Shared Autonomycited by 60
  36. 2024Learning Temporal Distances: Contrastive Successor Features Can Provide a Metric Structure for Decision-Makingcited by 56
  37. 2019Human few-shot learning of compositional instructionscited by 47
  38. 2016Learning to Act by Predicting the Futurecited by 36

50 most cited of 65 All 65 (JSON)

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 6; 2026: 8.

14 papers.

Sort:

By area:

  1. 2025Vision-Language-Action Models for Robotics: A Review Towards Real-World Applicationscited by 84
  2. 2025Vision Language Action Models in Robotic Manipulation: A Systematic Reviewcited by 38
  3. 2025Dual Goal Representationscited by 16
  4. 2025Self-Predictive Representations for Combinatorial Generalization in Behavioral Cloningcited by 12
  5. 2025Multistep Quasimetric Learning for Scalable Goal-conditioned Reinforcement Learningcited by 8
  6. 2025An Anatomy of Vision-Language-Action Models: From Modules to Milestones and Challengescited by 4
  7. 2026Three Steps at a Time: Learning Representations from Action Sequences in Contrastive RLcited by 3
  8. 2026Temporal Representations for Exploration: Learning Complex Exploratory Behavior without Extrinsic Rewardscited by 2
  9. 2026Action-Sufficient Goal Representationscited by 1
  10. 2026Mollified Value Learningcited by 1
  11. 2026MSPR: Multi-scale Predictive Representations for Goal-conditioned Reinforcement Learningcited by 1
  12. 2026Grounded Semantic Re-Binding for Robust Instruction Generalization in Vision-Language-Action Modelscited by 0
2 more
  1. 2026Do Better Goal Representations Improve Goal-Conditioned Reinforcement Learning?cited by 0
  2. 2026Imagining Goals with Geometric Horizon Models for Policy Generalizationcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Compose by Focus: Scene Graph-based Atomic Skills
  2. 2025Iterative Compositional Data Generation for Robot Control
  3. 2022Policy Architectures for Compositional Generalization in Control
  4. 2025RoboHiMan: A Hierarchical Evaluation Paradigm for Compositional Generalization in Long-Horizon Manipulation
  5. 2019Plan Arithmetic: Compositional Plan Vectors for Multi-Task Control
  6. 2026Emergent Compositional Skills in Mixture-of-Experts VLAs
  7. 2024RoboDreamer: Learning Compositional World Models for Robot Imagination
  8. 2025Learning Compositional Behaviors from Demonstration and Language