The paper introduces the Agent-Editing World Model (AEWM), a new approach that models how reasoning and actions influence future task progress instead of simulating tool responses. AEWM includes an Action Judge that classifies decisions as Critical, Exploratory, or Noisy, and a State Revision mechanism that edits noisy reasoning–action continuations from the same observed history. The integrated system, EditAct, directly updates the underlying state during real execution, leading to significant performance gains across multiple benchmarks and agent backbones.
By Shuang Sun, Guoxin Chen, Fanzhe Meng, Jia Deng, Huatong Song, Jinhao Jiang, Wayne Xin Zhao, Hongteng Xu, Ji-Rong Wen
arXiv:2609.24308v1 Announce Type: new
Abstract: Evaluating world models requires assessing both the quality of the worlds they generate and their consistency and responsiveness under exploration, int...
By Zhiqi Bai, Junai Cai, Yixin Chen, Jingrun Du, Tao Feng, Wei Gong, Siyuan Huang, Xiao Lin, Jiaheng Liu, Jun Luo, Yongzhe Lyu, Liya Ma, Zenan Meng, Lin Qu, Wenbo Su, Jiaming Wang, Qinghe Wang, Shaofei Wang, Yanghai Wang, Zequn Wang, Ziming Wang, Hu Wei, Jiangtao Wu, Ruiqi Wu, Jiaxin Xie, Yuchi Xu, Ze Xu, Chengting Yu, Liangyu Yuan, Gang Zeng, Yawen Zeng, Xingyao Zhang, Zizheng Zhang, Bo Zheng, Jiancheng Zhu, Song-Chun Zhu
arXiv:2610.28827v1 Announce Type: new
Abstract: We present WorldStream, an open-ended 3D world model with three simultaneous capabilities: it extends playable terrain beyond any pre-specified map, ac...
By Xiaoxiao Ma, Weiyang Jin, Yanwen Ma, Haoran Li, Zhixiang Wei, Songchun Zhang, Yaowei Li, Junhao Zhuang, Zeyue Xue, Yuming Li, Zheming Fu, Haoyang Huang, Nan Duan, Feng Zhao
Evaluating world models requires assessing both the quality of the worlds they generate and their consistency and responsiveness under exploration, interaction, and modification. We introduce HappyWor...
MultiWorldBench is a diagnostic Minecraft benchmark designed to test whether independently controlled views can describe a single shared world. It contains 495 case configurations across seven task suites and ten capabilities, such as independent control, cross-view motion, shared-state synchronization, persistence, structural reasoning, concurrent interaction, and delayed revisit. The benchmark evaluates systems like Solaris, Gamma-World, and MineWorld, revealing that none achieve high scores in state persistence, structural consistency, spatial reasoning, or building-identity preservation, and that Gamma-World and Solaris perform best in different subsets of capabilities.
By Zhangbo Xu, Ruoxi Zhang, Rui Hu, Yisong Wang
WorldAct is a framework that transforms static, monolithic 3D worlds generated by systems like Marble into editable, interaction‑ready scenes. It employs a multimodal agent to decompose the scene, identify actionable objects, reconstruct object‑level meshes, and inpaint the remaining background. The resulting scenes allow object‑level editing, collision‑aware manipulation, and embodied task execution while maintaining global coherence.
By Jichen Hu, Jiawei Guo, Jiazhong Cen, Chen Yang, Sikuang Li, Wei Shen
Code4Scene is a benchmark that evaluates coding agents on constructing and editing 3D scenes in Unreal Engine. It tests agents on two tasks: construction, where they must build a scene from open‑ended language, and editing, where they must recover a target scene from reference images while preserving everything else. The benchmark measures task fulfillment, artifact integrity, and physical validity, revealing that construction and editing performance are correlated but not interchangeable, with agents struggling most with spatial composition and precise edits.
By Xiaokang Ye, Siddhant Hitesh Mantri, Zimeng Chen, Edward Zhang, Zhaoxu Zheng, Yuanheng Li, Yizhao Chen, Tianyang Huang, Lianhui Qin
arXiv:2608. 11216v1 Announce Type: new Abstract: World modeling is an unsettled field: architectures, training objectives, and state representations interact in complex ways, and no single recipe dominates across environments.
By Marjan Moodi, Xuankang Zhu, Fernando De Mesentier Silva, Harold Chaput, Mohammad Reza Taesiri
The paper introduces RIPPLE, a method for adapting workflow-synthesizing agents through prompt-policy editing without retraining the underlying model. RIPPLE diagnoses failed execution trajectories, maps failures to specific policy segments, and restricts edits to those segments. It then evaluates candidate edits in isolation and replays only those that remain safe after composition, achieving up to a 23.1% improvement in validation success on a synthetic benchmark and positive gains on additional language‑model backbones.
By Manqing Mao, Hong Wang, Samson Koelle, Jie Yuan, Zhuoer Wang, James Feng, Yanjun Lin, Daniel Edmiston, Nikki Lijing Kuang, Zhecheng Sheng, Wei Niu
The paper investigates when generative models can function as simulators by examining their editability—whether altering a model’s latent representation yields outputs equivalent to editing a simulator’s state. The authors learn mappings between latent and simulator states using two approaches: decoding probes and an inverse map, and evaluate them in two environments, Othello and Rayworld. Results show that success depends on the environment: in Othello, editability degrades smoothly with simplification, while in Rayworld probes fail but an inverse map still aligns the model with the simulator’s state.
By Sevan Brodjian, Michael Hobley, Pietro Perona
arXiv:2609.36843v1 Announce Type: new
Abstract: Interpretability methods such as probes, activation patches and learned editors are designed to reveal or modify a model's current computation. World m...
By Junchi Yao, Ziyi Wang, Youling Huang, Lijie Hu
arXiv:2608. 09298v1 Announce Type: cross Abstract: Action-conditioned world models (ACWMs) promise to provide embodied AI with scalable predictive simulators for planning, policy evaluation, and data generation.
By Peterson Co, Sicheng Hu, Chunxuan Jiao, Hongyang Cheng, Yulin Luo, Yijie Xu, Sixiang Chen, Zhongxia Zhao, Zihao Wang, DaFeng Chi, Peidong Liu, YuTong Chen, Henghua Liu, Zhihao Yuan, Huizhu Jia, Yuzheng Zhuang, Tianle Zhang, Liang Lin, Huajie Tan, Shanghang Zhang