ActionSplice: In-Flight Action Editing for Interactive World Models

Hugging Face Daily Papers Papers

Summary

ActionSplice introduces counterfactual state transport to splice revised actions into chunk-autoregressive video world models without replaying completed evaluations, improving fidelity and inference speed.

Chunk-autoregressive video world models typically condition each generated chunk on one action. An action received during sampling must therefore wait for the next chunk, condition future solver evaluations on a state produced under the previous action, or trigger rollback that repeats completed evaluations. We introduce ActionSplice, an inference framework that formulates this problem as Counterfactual State Transport (CST). A lightweight corrector transports the interrupted backbone-native representation toward the matched 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. The retargeting variant CST*{R} updates the entire active chunk, while the temporal-splicing variant CST*{T} preserves a temporal prefix and updates only the suffix. Across minWM-Wan Action2V and HY-WM1.5, CST*{R} reduces rollback-relative LPIPS by 61.5% and 75.9% relative to direct condition swapping. CST*{T} reduces suffix LPIPS by 56.1% and 77.5%, respectively, while providing 2.73times and 1.69times pixel-ready speedups over waiting. Under the HY-WorldPlay protocol, CST_{R} obtains a PSNR of 25.66 dB, an SSIM of 0.6902, and an LPIPS of 0.1337 against the original rollout.
Original Article
View Cached Full Text

Cached at: 09/14/26, 02:35 PM

Paper page - ActionSplice: In-Flight Action Editing for Interactive World Models

Source: https://huggingface.co/papers/2609.08230 Published on Sep 8

·

Submitted byhttps://huggingface.co/PardisTaghavi

pardison Sep 14

Abstract

ActionSplice introduces counterfactual state transport to splice revised actions into chunk-autoregressive video world models without replaying completed evaluations, improving fidelity and speed.

Chunk-autoregressive video world modelstypically condition each generated chunk on one action. An action received during sampling must therefore wait for the next chunk, condition future solver evaluations on a state produced under the previous action, or trigger rollback that repeats completed evaluations. We introduceActionSplice, an inference framework that formulates this problem asCounterfactual State Transport(CST). A lightweightcorrectortransports the interruptedbackbone-native representationtoward the matched 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. The retargeting variant CST*{R} updates the entire active chunk, while the temporal-splicing variant CST*{T} preserves a temporal prefix and updates only the suffix. Across minWM-Wan Action2V and HY-WM1.5, CST*{R} reduces rollback-relativeLPIPSby 61.5% and 75.9% relative to direct condition swapping. CST*{T} reduces suffixLPIPSby 56.1% and 77.5%, respectively, while providing 2.73times and 1.69times pixel-ready speedups over waiting. Under the HY-WorldPlay protocol, CST_{R} obtains aPSNRof 25.66 dB, anSSIMof 0.6902, and anLPIPSof 0.1337 against the original rollout.

View arXiv pageView PDFProject pageGitHub3Add to collection

Get this paper in your agent:

hf papers read 2609\.08230

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2609.08230 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2609.08230 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2609.08230 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles

GameWAM: A World Action Model for Video Games

Hugging Face Daily Papers

GameWAM introduces the first unified world-action model for native video-game control, jointly predicting future visuals and executable actions using block-causal flow matching and mode-specific distributions.