mlsanityData to 7 Oct 2026

ActionSplice: In-Flight Action Editing for Interactive World Models

arXiv 2609.08230 ↗First posted 8 Sep 2026cs.CV cs.LGcited by 0

On the map: World models → Embodied and robot world models

Abstract (arXiv)

Chunk-autoregressive video world models typically generate each chunk under one action. When an action changes during sampling, waiting until the next chunk delays the response, while directly switching the conditioning leaves the intermediate solver state shaped by the previous action. Restarting sampling under the revised action avoids this mismatch but repeats completed computation. We introduce ActionSplice, which edits the interrupted state through Counterfactual State Transport (CST). A lightweight corrector moves the interrupted solver state toward the matched counterfactual solver state induced by the revised action at the same solver step. The world model and sampler remain frozen, and sampling resumes without replaying completed evaluations. CSTR retargets the entire active chunk, while CSTT preserves the temporal prefix at the intervention step and corrects only the suffix. On minWM--Wan Action2V and HY-WM1.5, we measure fidelity against matched rollback references. Relative to condition swapping, CSTR reduces LPIPS by 61.5% and 75.9%, and CSTT reduces suffix LPIPS by 56.1% and 77.5%, respectively.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Embodied and robot world models, one of 11 directions in World models. Builds on 24 of its 38 references that are in the map. First posted on arXiv 8 Sep 2026.

Builds on

24 of 38 references.

Closest ancestors · 1

Same direction, newest first

  1. 2026C3ache: Accelerating World Action Models with Cross Inference Chunk Cache8 Jun

Other references · 23

  1. 2021SDEdit: Guided Image Synthesis and Editing with Stochastic Differential Equationscited by 1,903
  2. 2024CameraCtrl: Enabling Camera Control for Text-to-Video Generationcited by 448
  3. 2025LongLive: Real-time Interactive Long Video Generationcited by 237
  4. 2025Sparse VideoGen: Accelerating Video Diffusion Transformers with Spatial-Temporal Sparsitycited by 204
  5. 2025Matrix-game 2.0: An open-source real-time and streaming interactive world modelcited by 174
  6. 2026Causal Forcing: Autoregressive Diffusion Distillation Done Right for High-Quality Real-Time Interactive Video Generationcited by 159
  7. 2025WorldPlay: Towards Long-Term Geometric Consistency for Real-Time Interactive World Modelingcited by 152
  8. 2026Matrix-Game 3.0: Real-Time and Streaming Interactive World Model with Long-Horizon Memorycited by 86
  9. 2025LongCat-Video Technical Reportcited by 73
  10. 2026Causal Forcing++: Scalable Few-Step Autoregressive Diffusion Distillation for Real-Time Interactive Video Generationcited by 44
  11. 2026Causal-rCM: A Unified Teacher-Forcing and Self-Forcing Open Recipe for Autoregressive Diffusion Distillation in Streaming Video Generation and Interactive World Modelscited by 23
  12. 2026minWM: A Full-Stack Open-Source Framework for Real-Time Interactive Video World Modelscited by 19
11 more
  1. 2026Anchor Forcing: Anchor Memory and Tri-Region RoPE for Interactive Streaming Video Diffusioncited by 14
  2. 2026Echo-Forcing: A Scene Memory Framework for Interactive Long Video Generationcited by 8
  3. 2026BiWM: Advancing Open-Source Interactive Video World Models with Bidirectional Autoregressioncited by 5
  4. 2026NaviDriveVLM: Decoupling High-Level Reasoning and Motion Planning for Autonomous Drivingcited by 4
  5. 2026X-Cache: Cross-Chunk Block Caching for Few-Step Autoregressive World Models Inferencecited by 4
  6. 2026Diffusion ReRoll: Revisable Denoising for Robotic Sequential Predictioncited by 4
  7. 2026Beyond Few-Step Inference: Accelerating Video Diffusion Transformer Model Serving with Inter-Request Caching Reusecited by 3
  8. 2026Light Interaction: Training-Free Inference Acceleration for Interactive Video World Modelscited by 2
  9. 2026Visko Orbis 1.0: A Live Model for Real-Time Interactive Long Video Generationcited by 2
  10. 2026Delta Forcing: Trust Region Steering for Interactive Autoregressive Video Generationcited by 1
  11. 2026Partition the Support, Reconstruct the Residual: Training-Free Sparse Attention for Video Generation and World Modelscited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026ReSync: Re-Aligning the Two Clocks of Asynchronous World-Action Models
  2. 2026Measuring the Stability Assumption Behind Action Chunking
  3. 2026FBFM: A Training-Free Asynchronous Feedback Mechanism for Flow-Matching in World-Action Models Execution
  4. 2026Twin Rollouts: Noise-Coupled Counterfactual Branching in Interactive Video World Models
  5. 2026SplineWAM: Adaptive Action Horizons for World Action Models via B-Spline Representations
  6. 2026Staircase Policy: Streaming Inference for World-Action Models with Large Action Chunks
  7. 2026WorldCycle: Self-Verifiable Reinforcement Learning for Long-Horizon Video World Models
  8. 2026ActWorld: From Explorable to Interactive World Model via Action-Aware Memory