mlsanityData to 14 Oct 2026

Compositional Transduction with Latent Analogies for Offline Goal-Conditioned Reinforcement Learning

arXiv 2605.20609 ↗First posted 20 May 2026cs.LGcited by 1

On the map: Offline RL and imitation learning from data → Diffusion and flow policies for offline RL

Abstract (arXiv)

Compositional generalization is essential for reaching unseen goals under novel contextual variations in offline goal-conditioned reinforcement learning (GCRL), where a generalist goal-reaching agent must be learned from limited data. Most prior approaches pursue this via trajectory stitching over temporally contiguous segments, which limits composing behaviors across varying contexts. To overcome this limitation, we formalize analogy transduction as synthesizing new plans by composing task-endogenous analogies with given contexts and propose a novel analogy representation tailored for it. Grounded in our theory, this analogy representation captures what changes under optimal task execution, remains invariant to contextual variations, and is sufficient for optimal goal reaching. We further contend that generalization to unseen analogy-context pairs is a practical obstacle in analogy transduction, and introduce a new approach for offline GCRL that enables analogy transduction beyond seen pairs to unseen combinations. We empirically demonstrate the effectiveness of our approach on OGBench manipulation environments, substantially outperforming prior methods that do not perform analogy transduction. Project page: https://rllab-snu.github.io/projects/CTA/

Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Placed in Diffusion and flow policies for offline RL, one of 13 directions in Offline RL and imitation learning from data. Builds on 62 of its 81 references that are in the map. First posted on arXiv 20 May 2026.

Builds on

62 of 81 references.

Closest ancestors · 8

Same direction, newest first

  1. 2025Offline Goal-conditioned Reinforcement Learning with Quasimetric Representations24 Sep
  2. 2025Physics-informed Value Learner for Offline Goal-Conditioned Reinforcement Learning8 Sep
  3. 2024OGBench: Benchmarking Offline Goal-Conditioned RL26 Oct
  4. 2024Stitching Sub-Trajectories with Conditional Diffusion Model for Goal-Conditioned Offline RL11 Feb
  5. 2023Diffused Task-Agnostic Milestone Planner6 Dec
  6. 2023SMORE: Score Models for Offline Goal-Conditioned Reinforcement Learning3 Nov
  7. 2022Offline Reinforcement Learning via High-Fidelity Generative Behavior Modeling29 Sep
  8. 2022How Far I'll Go: Offline Goal-Conditioned Reinforcement Learning via f-Advantage Regression7 Jun

Other references · 54

  1. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
  2. 2018A Style-Based Generator Architecture for Generative Adversarial Networkscited by 6,013
  3. 2016Layer Normalizationcited by 5,669
  4. 2018Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actorcited by 4,885
  5. 2013Efficient Estimation of Word Representations in Vector Spacecited by 4,375
  6. 2016Gaussian Error Linear Units (GELUs)cited by 3,502
  7. 2013Playing Atari with Deep Reinforcement Learningcited by 3,274
  8. 2020Offline Reinforcement Learning: Tutorial, Review, and Perspectives on Open Problemscited by 1,920
  9. 2021Offline Reinforcement Learning with Implicit Q-Learningcited by 1,293
  10. 2017Hindsight Experience Replaycited by 1,136
  11. 2022Planning with Diffusion for Flexible Behavior Synthesiscited by 1,061
  12. 2021A Minimalist Approach to Offline Reinforcement Learningcited by 899
38 more
  1. 2019Advantage-Weighted Regression: Simple and Scalable Off-Policy Reinforcement Learningcited by 782
  2. 2020Learning Invariant Representations for Reinforcement Learning without Reconstructioncited by 436
  3. 2022VIP: Towards Universal Visual Reward and Representation via Value-Implicit Pre-Trainingcited by 395
  4. 2022Contrastive Learning as Goal-Conditioned Reinforcement Learningcited by 253
  5. 2019Provably efficient RL with Rich Observations via Latent State Decodingcited by 169
  6. 2023HIQL: Offline Goal-Conditioned RL with Latent States as Actionscited by 141
  7. 2019Scalable methods for computing state similarity in deterministic Markov Decision Processescited by 133
  8. 2013Learning by Transductioncited by 121
  9. 2023Optimal Goal-Reaching Reinforcement Learning via Quasimetric Learningcited by 121
  10. 2024Foundation Policies with Hilbert Representationscited by 94
  11. 2023METRA: Scalable Unsupervised RL with Metric-Aware Abstractioncited by 86
  12. 2023For SALE: State-Action Representation Learning for Deep Reinforcement Learningcited by 75
  13. 2023Reinforcement Learning from Passive Data via Latent Intentionscited by 68
  14. 2023Elastic Decision Transformercited by 67
  15. 2023Compositional Generalization from First Principlescited by 64
  16. 2020Quantifying Differences in Reward Functionscited by 63
  17. 2024Learning Temporal Distances: Contrastive Successor Features Can Provide a Metric Structure for Decision-Makingcited by 56
  18. 2018Learning Actionable Representations with Goal-Conditioned Policiescited by 54
  19. 2022Guaranteed Discovery of Control-Endogenous Latent States with Multi-Step Inverse Modelscited by 45
  20. 2025Generative Trajectory Stitching through Diffusion Compositioncited by 41
  21. 2022Bisimulation Makes Analogies in Goal-Conditioned Reinforcement Learningcited by 38
  22. 2023Contrastive Difference Predictive Codingcited by 36
  23. 2024DiffStitch: Boosting Offline Reinforcement Learning with Diffusion-based Trajectory Stitchingcited by 36
  24. 2018Grasp2Vec: Learning Object Representations from Self-Supervised Graspingcited by 27
  25. 2024Closing the Gap between TD Learning and Supervised Learning -- A Generalisation Point of Viewcited by 27
  26. 2024Reinformer: Max-Return Sequence Modeling for Offline RLcited by 25
  27. 2022Improved Representation of Asymmetrical Distances with Interval Quasimetric Embeddingscited by 24
  28. 2023Learning to Extrapolate: A Transductive Approachcited by 19
  29. 2022Dynamics-Aware Comparison of Learned Reward Functionscited by 17
  30. 2025Horizon Generalization in Reinforcement Learningcited by 17
  31. 2025Dual Goal Representationscited by 16
  32. 2022On the Learning and Learnability of Quasimetricscited by 15
  33. 2023Multi-task Hierarchical Adversarial Inverse Reinforcement Learningcited by 15
  34. 2023STARC: A General Framework For Quantifying Differences Between Reward Functionscited by 15
  35. 2022Metric Residual Networks for Sample Efficient Goal-Conditioned Reinforcement Learningcited by 13
  36. 2023Free from Bellman Completeness: Trajectory Stitching via Model-based Return-conditioned Supervised Learningcited by 13
  37. 2019Plan Arithmetic: Compositional Plan Vectors for Multi-Task Controlcited by 11
  38. 2024Bisimulation Metrics are Optimal Transport Distances, and Can be Computed Efficientlycited by 6

50 most cited of 54 All 54 (JSON)

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026HorizonFlow: Variable-Length Planning for Offline Goal-Conditioned RLcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Modeling Unseen Environments with Language-guided Composable Causal Components in Reinforcement Learning
  2. 2026Compositional Reasoning in Language Models under Reinforcement Learning Post-Training
  3. 2026Learning Multiple Timescales for Goal-Conditioned Reinforcement Learning
  4. 2022Generalization with Lossy Affordances: Leveraging Broad Offline Data for Learning Visuomotor Tasks
  5. 2026Understanding Goal Generalisation in Sequential Reinforcement Learning
  6. 2020Learning to Generalize for Sequential Decision Making
  7. 2022Policy Architectures for Compositional Generalization in Control
  8. 2024Generalization of Compositional Tasks with Logical Specification via Implicit Planning