mlsanityData to 7 Oct 2026

Does Continual Imitation Learning Remain Grounded? A Language-Perturbed Benchmark for Robotic Task Retention

arXiv 2610.00542 ↗First posted 30 Sep 2026cs.ROcited by 0

On the map: Offline RL and imitation learning from data → Language-conditioned policies and VLAs

Abstract (arXiv)

Continual imitation learning evaluates whether a robot can learn new knowledge without forgetting previously learned skills. However, retaining task performance does not ensure the behavior remains grounded in language because policies may rely on scene cues, object associations, or memorized task structure. We introduce a benchmark protocol to study how language-guided behavior changes as robotic policies learn successive tasks. We construct meaning-preserving and meaning-changing instruction variants for the Goal, Spatial, Object, and Long suites of LIBERO. Policy experiments focus on LIBERO-Goal, evaluating Original and Paraphrase instructions after each continual-learning stage. We compare representative continual imitation learning methods under their original assumptions while separating task competence from language sensitivity. The proposed diagnostics complement standard learning and forgetting metrics by measuring semantic robustness, goal adaptation, and language sensitivity. Results show that strong continual-learning performance does not always translate to reliable language grounding, and our diagnostics help determine whether retained skills remain correctly guided by their instructions. Additional materials are available at https://sites.google.com/view/stillgrounded

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Language-conditioned policies and VLAs, one of 13 directions in Offline RL and imitation learning from data. Builds on 28 of its 30 references that are in the map. First posted on arXiv 30 Sep 2026.

Builds on

28 of 30 references.

References · 28

  1. 2016Overcoming catastrophic forgetting in neural networkscited by 4,718
  2. 2017Gradient Episodic Memory for Continual Learningcited by 1,871
  3. 2023LIBERO: Benchmarking Knowledge Transfer for Lifelong Robot Learningcited by 1,338
  4. 2016Progressive Neural Networkscited by 1,332
  5. 2017Memory Aware Synapses: Learning what (not) to forgetcited by 1,094
  6. 2018Experience Replay for Continual Learningcited by 1,074
  7. 2021Learning to Prompt for Continual Learningcited by 938
  8. 2017PackNet: Adding Multiple Tasks to a Single Network by Iterative Pruningcited by 882
  9. 2022DualPrompt: Complementary Prompting for Rehearsal-free Continual Learningcited by 635
  10. 2021CLIPort: What and Where Pathways for Robotic Manipulationcited by 617
  11. 2022BC-Z: Zero-Shot Task Generalization with Robotic Imitation Learningcited by 517
  12. 2022CODA-Prompt: COntinual Decomposed Attention-based Prompting for Rehearsal-Free Continual Learningcited by 446
16 more
  1. 2016Expert Gate: Lifelong Learning with a Network of Expertscited by 369
  2. 2017PathNet: Evolution Channels Gradient Descent in Super Neural Networkscited by 362
  3. 2025LIBERO-Plus: In-depth Robustness Analysis of Vision-Language-Action Modelscited by 278
  4. 2020Towards Continual Reinforcement Learning: A Review and Perspectivescited by 268
  5. 2025LIBERO-PRO: Towards Robust and Fair Evaluation of Vision-Language-Action Models Beyond Memorizationcited by 166
  6. 2024Continual Learning with Pre-Trained Models: A Surveycited by 151
  7. 2021Continual World: A Robotic Benchmark For Continual Reinforcement Learningcited by 112
  8. 2023A Definition of Continual Reinforcement Learningcited by 98
  9. 2023TAIL: Task-specific Adapters for Imitation Learning with Large Pretrained Modelscited by 48
  10. 2023Learning to Modulate pre-trained Models in RLcited by 28
  11. 2024Incremental Learning of Retrievable Skills For Efficient Continual Task Adaptationcited by 19
  12. 2023Online Class-Incremental Learning For Real-World Food Image Classificationcited by 17
  13. 2024DELTA: Decoupling Long-Tailed Online Continual Learningcited by 8
  14. 2025Dynamic Mixture of Progressive Parameter-Efficient Expert Library for Lifelong Robot Learningcited by 8
  15. 2024Contrast Sets for Evaluating Language-Guided Robot Policiescited by 7
  16. 2026Can You Hear, Localize, and Segment Continually? An Exemplar-Free Continual Learning Benchmark for Audio-Visual Segmentationcited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026Can Vision-Language-Action Models Learn from Real-World Data Continually without Forgetting?
  2. 2026Pretrained Vision-Language-Action Models are Surprisingly Resistant to Forgetting in Continual Learning
  3. 2026Lifelong Imitation Learning with Multimodal Latent Replay and Incremental Adjustment
  4. 2024Rethinking Mutual Information for Language Conditioned Skill Discovery on Imitation Learning
  5. 2023Language-Conditioned Imitation Learning with Base Skill Priors under Unstructured Data
  6. 2025SPECI: Skill Prompts based Hierarchical Continual Imitation Learning for Robot Manipulation
  7. 2020Language Conditioned Imitation Learning over Unstructured Data
  8. 2026Benchmarking World Models for Continual Learning on Compositional Tasks