mlsanityData to 13 Oct 2026

Generalizing to New Tasks via One-Shot Compositional Subgoals

arXiv 2205.07716 ↗First posted 16 May 2022cs.LG cs.ROcited by 1

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

Abstract (arXiv)

The ability to generalize to previously unseen tasks with little to no supervision is a key challenge in modern machine learning research. It is also a cornerstone of a future "General AI". Any artificially intelligent agent deployed in a real world application, must adapt on the fly to unknown environments. Researchers often rely on reinforcement and imitation learning to provide online adaptation to new tasks, through trial and error learning. However, this can be challenging for complex tasks which require many timesteps or large numbers of subtasks to complete. These "long horizon" tasks suffer from sample inefficiency and can require extremely long training times before the agent can learn to perform the necessary longterm planning. In this work, we introduce CASE which attempts to address these issues by training an Imitation Learning agent using adaptive "near future" subgoals. These subgoals are recalculated at each step using compositional arithmetic in a learned latent representation space. In addition to improving learning efficiency for standard long-term tasks, this approach also makes it possible to perform one-shot generalization to previously unseen tasks, given only a single reference trajectory for the task in a different environment. Our experiments show that the proposed approach consistently outperforms the previous state-of-the-art compositional Imitation Learning approach by 30%.

Key facts. Cited by 1 paper in the map (data through 13 Oct 2026). Placed in Robot manipulation imitation learning, one of 13 directions in Offline RL and imitation learning from data. Builds on 24 of its 40 references that are in the map. First posted on arXiv 16 May 2022.

Builds on

24 of 40 references.

Closest ancestors · 2

Same direction, newest first

  1. 2022SKILL-IL: Disentangling Skill and Knowledge in Multitask Imitation Learning6 May
  2. 2019Relay Policy Learning: Solving Long-Horizon Tasks via Imitation and Reinforcement Learning25 Oct

Other references · 22

  1. 2013Distributed Representations of Words and Phrases and their Compositionalitycited by 3,467
  2. 2016Generative Adversarial Imitation Learningcited by 1,577
  3. 2018Data-Efficient Hierarchical Reinforcement Learningcited by 442
  4. 2019Learning Latent Plans from Playcited by 313
  5. 2018Semi-parametric Topological Memory for Navigationcited by 204
  6. 2018Multi-Agent Generative Adversarial Imitation Learningcited by 99
  7. 2017Zero-Shot Task Generalization with Multi-Task Deep Reinforcement Learningcited by 97
  8. 2017Deeply AggreVaTeD: Differentiable Imitation Learning for Sequential Predictioncited by 93
  9. 2019IRIS: Implicit Reinforcement without Interaction at Scale for Learning Control from Offline Robot Manipulation Datacited by 85
  10. 2017Hierarchical and Interpretable Skill Acquisition in Multi-task Reinforcement Learningcited by 77
  11. 2018TACO: Learning Task Decomposition via Temporal Alignment for Controlcited by 61
  12. 2020Generating Adjacency-Constrained Subgoals in Hierarchical Reinforcement Learningcited by 57
10 more
  1. 2021Landmark-Guided Subgoal Generation in Hierarchical Reinforcement Learningcited by 47
  2. 2018Truncated Horizon Policy Search: Combining Reinforcement Learning & Imitation Learningcited by 39
  3. 2014Interactive Policy Learning through Confidence-Based Autonomycited by 29
  4. 2018Fast Policy Learning through Imitation and Reinforcementcited by 27
  5. 2019Learning from Trajectories via Subgoal Discoverycited by 27
  6. 2020Hallucinative Topological Memory for Zero-Shot Visual Planningcited by 25
  7. 2021Subgoal Search For Complex Reasoning Taskscited by 21
  8. 2017Intention-Net: Integrating Planning and Deep Learning for Goal-Directed Autonomous Navigationcited by 19
  9. 2021Active Hierarchical Exploration with Stable Subgoal Representation Learningcited by 17
  10. 2021Robot in a China Shop: Using Reinforcement Learning for Location-Specific Navigation Behaviourcited by 2

Built on by

Cited by year (2026 to 13 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026Multi-Domain Learning with Global Expert Mappingcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2019Plan Arithmetic: Compositional Plan Vectors for Multi-Task Control
  2. 2022Policy Architectures for Compositional Generalization in Control
  3. 2021Procedural Generalization by Planning with Self-Supervised World Models
  4. 2020Learning to Generalize Across Long-Horizon Tasks from Human Demonstrations
  5. 2026Compositional Transduction with Latent Analogies for Offline Goal-Conditioned Reinforcement Learning
  6. 2023Task Aware Dreamer for Task Generalization in Reinforcement Learning
  7. 2021Towards More Generalizable One-shot Visual Imitation Learning
  8. 2023Improving Generalization in Game Agents with Data Augmentation in Imitation Learning