Related papers: ActionSplice: In-Flight Action Editing for Interactive World Models

ActionSplice: In-Flight Action Editing for Interactive World Models

URL: http://arxiv.org/abs/2609.08230v1
Date: Tue, 08 Sep 2026 04:21:36 GMT
Title: ActionSplice: In-Flight Action Editing for Interactive World Models
Abstract summary: Chunk-autoregressive video world models typically condition each generated chunk on one action.<n>We introduce Actionplice, an inference framework that formulates this problem as Counterfactual State Transport (CST)<n>A temporal-splicing variant $mathrmCST*R$ preserves a temporal prefix and updates only the suffix.<n>$mathrmCST*T$ reduces suffix LPIPS by 56.1% and 77.5%, respectively, while providing $2.73times$ and $1.69times$ pixel-ready speedups over waiting.
Score: 4.400011068855375
License: http://creativecommons.org/licenses/by/4.0/
Abstract: 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 $\mathrm{CST}*{R}$ updates the entire active chunk, while the temporal-splicing variant $\mathrm{CST}*{T}$ preserves a temporal prefix and updates only the suffix. Across minWM-Wan Action2V and HY-WM1.5, $\mathrm{CST}*{R}$ reduces rollback-relative LPIPS by 61.5% and 75.9% relative to direct condition swapping. $\mathrm{CST}*{T}$ reduces suffix LPIPS by 56.1% and 77.5%, respectively, while providing $2.73\times$ and $1.69\times$ pixel-ready speedups over waiting. Under the HY-WorldPlay protocol, $\mathrm{CST}_{R}$ obtains a PSNR of 25.66 dB, an SSIM of 0.6902, and an LPIPS of 0.1337 against the original rollout.

Related papers

Measuring the Value of World-Model Updates: A Counterfactual Utility Protocol for Continual Adaptation [0.0]
We introduce the fork ledger, which branches a deployment stream at pre-registered decision points into matched update and hold continuations under common random numbers.<n>It evaluates both continuations on the same episodes and records $R = R_mathrmupdate - R_mathrmhold$.
arXiv Detail & Related papers (2026-09-10T01:22:57Z)
Hidden Boundary Motion in Transformer Optimization: Function-Space Orthogonalization of Affine Weight and Bias Updates [0.0]
We call this hidden contribution emphboundary motion and decompose each update into a emphshape component and a shared emphboundary component.<n>In a single-seed experiment, SBO-AdamW raises validation accuracy from 81.68% to 85.81% and validation-selected test accuracy from 78.73% to 82.73%.
arXiv Detail & Related papers (2026-07-24T21:51:43Z)
Polymorphism Is Rotation: Operational Mechanistic Interpretability from a Two-Layer Transformer to Pythia-70m [0.0]
Independently trained transformers compute the same function in residual-stream bases that differ by a uniform random rotation.<n>We call this phenomenon polymorphism: same function, mutually unintelligible interior coordinates.<n>The phenomenon is invisible to the standard SAE metric.
arXiv Detail & Related papers (2026-05-23T13:37:59Z)
Minimax Optimal Variance-Aware Regret Bounds for Multinomial Logistic MDPs [56.28491566735463]
Existing algorithms for episodic Markov Decision Processes yield a regret of $smashtildeO(dH2sqrtT)$.<n>We propose an algorithm achieving a regret of $smashtildeO(dH2bar_TsqrtT)$, which recovers the existing bound in the worst case and improves upon it for structured MDPs.
arXiv Detail & Related papers (2026-05-19T12:39:32Z)
Latent Phase-Shift Rollback: Inference-Time Error Correction via Residual Stream Monitoring and KV-Cache Steering [4.032680910442999]
Large language models commit unrecoverable reasoning errors mid-generation.<n>We introduce $textbfLatent Phase-Shift Rollback$ (LPSR)<n>At each generation step, we monitor the residual stream at a critical layer lcrit, detect abrupt directional reversals (phase shifts) via a cosine-similarity $+$ entropy dual gate.<n>No fine-tuning, gradient computation, or additional forward passes are required.
arXiv Detail & Related papers (2026-04-20T17:53:33Z)
S0 Tuning: Zero-Overhead Adaptation of Hybrid Recurrent-Attention Models [0.0]
tuning a single initial state matrix per recurrent layer with zero overhead outperforms LoRA by +10.8 pp (p wenSpider on HumanEval)<n> tuning method we call S0 tuning improves greedy pass@1 by +23.6 +/- 1.7 pp (10 seeds)<n>Cross-domain transfer is significant on MATH-500 (+4.8 pp, p = 0.00002, 8 seeds) and GSM8K = 0.0003, 10 seeds.
arXiv Detail & Related papers (2026-04-01T17:21:15Z)
Improved Algorithm for Adversarial Linear Mixture MDPs with Bandit Feedback and Unknown Transition [71.33787410075577]
We study reinforcement learning with linear function approximation, unknown transition, and adversarial losses. We propose a new algorithm that attains an $widetildeO(dsqrtHS3K + sqrtHSAK)$ regret with high probability.
arXiv Detail & Related papers (2024-03-07T15:03:50Z)
Near Sample-Optimal Reduction-based Policy Learning for Average Reward MDP [58.13930707612128]
This work considers the sample complexity of obtaining an $varepsilon$-optimal policy in an average reward Markov Decision Process (AMDP) We prove an upper bound of $widetilde O(H varepsilon-3 ln frac1delta)$ samples per state-action pair, where $H := sp(h*)$ is the span of bias of any optimal policy, $varepsilon$ is the accuracy and $delta$ is the failure probability.
arXiv Detail & Related papers (2022-12-01T15:57:58Z)
Reward-Mixing MDPs with a Few Latent Contexts are Learnable [75.17357040707347]
We consider episodic reinforcement learning in reward-mixing Markov decision processes (RMMDPs) Our goal is to learn a near-optimal policy that nearly maximizes the $H$ time-step cumulative rewards in such a model.
arXiv Detail & Related papers (2022-10-05T22:52:00Z)
Optimistic Posterior Sampling for Reinforcement Learning with Few Samples and Tight Guarantees [43.13918072870693]
We propose an optimistic posterior sampling algorithm for reinforcement learning (OPSRL) We guarantee a high-probability regret bound of order at most $widetildemathcalO(sqrtH3SAT)$ ignoring $textpolylog(HSAT)$ terms. Our bound matches the lower bound of order $Omega(sqrtH3SAT)$, thereby answering the open problems raised by Agrawal and Jia.
arXiv Detail & Related papers (2022-09-28T20:49:34Z)
Stochastic Shortest Path: Minimax, Parameter-Free and Towards Horizon-Free Regret [144.6358229217845]
We study the problem of learning in the shortest path (SSP) setting, where an agent seeks to minimize the expected cost accumulated before reaching a goal state. We design a novel model-based algorithm EB-SSP that carefully skews the empirical transitions and perturbs the empirical costs with an exploration bonus. We prove that EB-SSP achieves the minimax regret rate $widetildeO(B_star sqrtS A K)$, where $K$ is the number of episodes, $S$ is the number of states, $A$
arXiv Detail & Related papers (2021-04-22T17:20:48Z)
Sample Complexity of Asynchronous Q-Learning: Sharper Analysis and Variance Reduction [63.41789556777387]
Asynchronous Q-learning aims to learn the optimal action-value function (or Q-function) of a Markov decision process (MDP) We show that the number of samples needed to yield an entrywise $varepsilon$-accurate estimate of the Q-function is at most on the order of $frac1mu_min (1-gamma)5varepsilon2+ fract_mixmu_min (1-gamma)$ up to some logarithmic factor.
arXiv Detail & Related papers (2020-06-04T17:51:00Z)

This list is automatically generated from the titles and abstracts of the papers in this site.