mlsanityData to 7 Oct 2026

ACID: Action Consistency via Inverse Dynamics for Planning with World Models

arXiv 2607.02403 ↗First posted 2 Jul 2026cs.RO cs.AI cs.CVcited by 4

On the map: World models → Latent predictive world models

Abstract (arXiv)

Decision-time planning with action-conditioned world models has become a popular paradigm for embodied control. However, the standard planning cost judges a candidate solely by how close its predicted terminal state lies to the goal, leaving the realizability of the intermediate transitions unchecked--a predicted trajectory can look convincing while the environment rollout drifts away from it. In this paper, we propose ACID, a decision-time planning framework that introduces cycle action consistency: the action inferred backward from a predicted transition by an inverse dynamics model should recover the one that was conditioned on. We fold this per-step residual into the planning cost via a scale-invariant adaptive weight. Across four action-conditioned world models and eight tasks encompassing object manipulation and articulated control in simulation, visual navigation, and real-robot manipulation, ACID consistently improves planning and matches the baseline's accuracy with substantially less planning compute.

Key facts. Cited by 4 papers in the map (data through 7 Oct 2026). Placed in Latent predictive world models, one of 11 directions in World models. Builds on 35 of its 40 references that are in the map. First posted on arXiv 2 Jul 2026.

Builds on

35 of 40 references.

Closest ancestors · 3

Same direction, newest first

  1. 2026LeWorldModel: Stable End-to-End Joint-Embedding Predictive Architecture from Pixels13 Mar
  2. 2026Planning in 8 Tokens: A Compact Discrete Tokenizer for Latent World Model5 Mar
  3. 2024DINO-WM: World Models on Pre-trained Visual Features enable Zero-shot Planning7 Nov

Other references · 32

  1. 2018The Unreasonable Effectiveness of Deep Features as a Perceptual Metriccited by 11,029
  2. 2022Flow Matching for Generative Modelingcited by 5,426
  3. 2024π0: A Vision-Language-Action Flow Model for General Robot Controlcited by 2,297
  4. 2018World Modelscited by 1,444
  5. 2018DeepMind Control Suitecited by 935
  6. 2025V-JEPA 2: Self-Supervised Video Models Enable Understanding, Prediction and Planningcited by 738
  7. 2024Genie: Generative Interactive Environmentscited by 706
  8. 2024Diffusion Forcing: Next-token Prediction Meets Full-Sequence Diffusioncited by 634
  9. 2023GAIA-1: A Generative World Model for Autonomous Drivingcited by 618
  10. 2023Learning Universal Policies via Text-Guided Video Generationcited by 506
  11. 2023TD-MPC2: Scalable, Robust World Models for Continuous Controlcited by 482
  12. 2026World Action Models are Zero-shot Policiescited by 411
20 more
  1. 2022Video PreTraining (VPT): Learning to Act by Watching Unlabeled Online Videoscited by 338
  2. 2024Vista: A Generalizable Driving World Model with High Fidelity and Versatile Controllabilitycited by 337
  3. 2022Rectified Flow: A Marginal Preserving Approach to Optimal Transportcited by 277
  4. 2024Navigation World Modelscited by 262
  5. 2025LeJEPA: Provable and Scalable Self-Supervised Learning Without the Heuristicscited by 227
  6. 2025Unified World Models: Coupling Video and Action Diffusion for Pretraining on Large Robotic Datasetscited by 218
  7. 2025DreamGen: Unlocking Generalization in Robot Learning through Video World Modelscited by 168
  8. 2025History-Guided Video Diffusioncited by 165
  9. 2023Video Language Planningcited by 160
  10. 2021Probabilistic Embeddings for Cross-Modal Retrievalcited by 150
  11. 2026GigaWorld-Policy: An Efficient Action-Centered World--Action Modelcited by 129
  12. 2021Rapid Exploration for Open-World Navigation with Latent Goal Modelscited by 118
  13. 2019Polysemous Visual-Semantic Embedding for Cross-Modal Retrievalcited by 83
  14. 2025Learning from Reward-Free Offline Data: A Case for Planning with Latent Dynamics Modelscited by 77
  15. 2024AdaptiGraph: Material-Adaptive Graph-Based Neural Dynamics for Robotic Manipulationcited by 52
  16. 2024AVID: Adapting Video Diffusion Models to World Modelscited by 35
  17. 2022Improving Cross-Modal Retrieval with Set of Diverse Embeddingscited by 33
  18. 2023Inverse Dynamics Pretraining Learns Good Representations for Multitask Imitationcited by 30
  19. 2025Latent Diffusion Planning for Imitation Learningcited by 26
  20. 2025FreeAction: Training-Free Techniques for Enhanced Fidelity of Trajectory-to-Video Generationcited by 4

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 4.

4 papers.

  1. 2026Diagnosing JEPA World Models with Action-Conditioned Predictive Consistencycited by 2
  2. 2026Paired Exact-Reset Evaluation of a Prediction-Derived Medium-to-Full World-Model Cascadecited by 0
  3. 2026Beyond Visual Quality: A Study of Test-Time Planning with World Action Modelscited by 0
  4. 2026NeuronDiscover: Agent-in-Twin for Mechanistic Discovery in Neuronal Microenvironments with World Action Modelscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026Latent Geometry Beyond Search: Amortizing Planning in World Models
  2. 2025World Modeling Makes a Better Planner: Dual Preference Optimization for Embodied Task Planning
  3. 2026AdaReP:Adaptive Re-Planning under Model Mismatch for Neural World-Model Predictive Control
  4. 2026Toward Physically Grounded JEPA World Models for Goal-Conditioned Robotic Planning
  5. 2020Planning from Pixels using Inverse Dynamics Models
  6. 2026Representation World Model: Learning States, Transition and Executable Plans in Representation
  7. 2026Unifying Object-Centric World Models and Diffusion Policy: A Hierarchical Framework for Multi-Stage Robotic Tasks
  8. 2026Hydra: A Navigation World Action Model with Discrete Latent Planning and Continuous Flow-Matching Execution