ActionSplice: In-Flight Action Editing for Interactive World Models
ActionSplice enables in-flight action editing in chunk-autoregressive video world models via a lightweight corrector, avoiding rollback or waiting for the next chunk.
ActionSplice is an inference framework that formulates in-flight action editing for chunk-autoregressive video world models as Counterfactual State Transport (CST), where a lightweight corrector transports the interrupted backbone-native representation toward the matched state induced by the revised action. The world model and sampler remain frozen, and sampling resumes without replaying completed evaluations. Across minWM-Wan Action2V and HY-WM1.5, the retargeting variant CST-R reduces rollback-relative LPIPS by 61.5% and 75.9% versus direct condition swapping, while the temporal-splicing variant CST-T reduces suffix LPIPS by 56.1% and 77.5% with 2.73x and 1.69x pixel-ready speedups over waiting.
- Counterfactual State Transport edits actions mid-chunk without rollback
- World model and sampler remain completely frozen
- CST-R cuts rollback-relative LPIPS by 61.5% and 75.9% on two models
- Temporal-splicing variant delivers 2.73x and 1.69x speedups over waiting
From the original article
1 min read at huggingface.co
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.
By Pardis Taghavi, Tingyu Guo, Jonas Lossner · huggingface.co
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.