[ { "id": "P01", "alias": "3D-ViTac", "title": "3D-ViTac: Learning Fine-Grained Manipulation with Visuo-Tactile Sensing", "authors": [ "Huang, Binghao", "Wang, Yixuan", "Yang, Xinyi", "Luo, Yiyue", "Li, Yunzhu" ], "date": "2024/10/31", "category": "Policy and fusion", "origin": "Linked in seed", "priority": "Core", "depth": "Methods/results", "sensing": "Pressure array + 3D vision", "contribution": "Spatially aligns tactile measurements and visual geometry for fine manipulation.", "boundary": "Baseline for FlexiTac representations; array design and calibration must match the actual hardware.", "url": "https://arxiv.org/abs/2410.24091", "fulltext_url": "https://arxiv.org/html/2410.24091v1", "accessed": "2026-09-08" }, { "id": "P02", "alias": "VT-Refine", "title": "VT-Refine: Learning Bimanual Assembly with Visuo-Tactile Feedback via Simulation Fine-Tuning", "authors": [ "Huang, Binghao", "Xu, Jie", "Akinola, Iretiayo", "Yang, Wei", "Sundaralingam, Balakumar", "O'Flaherty, Rowland", "Fox, Dieter", "Wang, Xiaolong", "Mousavian, Arsalan", "Chao, Yu-Wei", "Li, Yunzhu" ], "date": "2025/10/16", "category": "Simulation and sensors", "origin": "Linked in seed", "priority": "Core", "depth": "Methods/results + code", "sensing": "Pressure array + vision", "contribution": "Uses real demonstrations and calibrated tactile simulation to fine-tune bimanual assembly policies.", "boundary": "Closest pressure-array transfer precedent; rigid CAD tasks do not establish general deformable or synthetic-only VLA learning.", "url": "https://arxiv.org/abs/2510.14930", "fulltext_url": "https://arxiv.org/html/2510.14930v1", "accessed": "2026-09-08" }, { "id": "P03", "alias": "Policy Consensus", "title": "Multi-Modal Manipulation via Multi-Modal Policy Consensus", "authors": [ "Chen, Haonan", "Xu, Jiaming", "Chen, Hongyu", "Hong, Kaiwen", "Huang, Binghao", "Liu, Chaoqi", "Mao, Jiayuan", "Li, Yunzhu", "Du, Yilun", "Driggs-Campbell, Katherine" ], "date": "2025/09/27", "category": "Policy and fusion", "origin": "Linked in seed", "priority": "Core", "depth": "Methods/results", "sensing": "Pressure array in real tasks; other modalities in RLBench", "contribution": "Combines modality-specific policies through agreement to improve manipulation.", "boundary": "Use as a strong fusion control; its RLBench results are not tactile experiments.", "url": "https://arxiv.org/abs/2509.23468", "fulltext_url": "https://arxiv.org/html/2509.23468v1", "accessed": "2026-09-08" }, { "id": "P04", "alias": "PhysTwin", "title": "PhysTwin: Physics-Informed Reconstruction and Simulation of Deformable Objects from Videos", "authors": [ "Jiang, Hanxiao", "Hsu, Hao-Yu", "Zhang, Kaifeng", "Yu, Hsin-Ni", "Wang, Shenlong", "Li, Yunzhu" ], "date": "2025/03/23", "category": "Simulation and sensors", "origin": "Linked in seed", "priority": "Supporting", "depth": "Methods/results", "sensing": "Video to deformable digital twin", "contribution": "Reconstructs deformable objects with a physics-based representation and visual appearance.", "boundary": "Useful later for deformable assets; visual reconstruction alone does not calibrate tactile forces.", "url": "https://arxiv.org/abs/2503.17973", "fulltext_url": "https://arxiv.org/html/2503.17973v1", "accessed": "2026-09-08" }, { "id": "P05", "alias": "FELT", "title": "FELT: Generating Tactile Signals from Vision for Visuo-Tactile Manipulation", "authors": [ "Li, Zinan", "Ling, Yiyang", "Gu, Yuming", "Huang, Binghao", "Liang, Chenhao", "Islam, Sharfin", "Bedri, Hisham", "Chirikjian, John", "Li, Yunzhu", "Nikolaidis, Stefanos", "Seita, Daniel" ], "date": "2026/07/22", "category": "Data and representations", "origin": "Linked in seed", "priority": "Core", "depth": "Methods/results", "sensing": "Vision-predicted pressure", "contribution": "Learns tactile signal generation from paired visual and tactile data for manipulation.", "boundary": "Include as a pseudo-touch baseline; inferred touch cannot supply independent measurements of visually hidden contact.", "url": "https://arxiv.org/abs/2607.20683", "fulltext_url": "https://arxiv.org/html/2607.20683v1", "accessed": "2026-09-08" }, { "id": "P06", "alias": "Beyond Teleoperation", "title": "Beyond Teleoperation: Enhancing VLA Robustness via Explicit Kinematic Retargeting of Human Demonstrations", "authors": [ "José António Nunes Andrade", "Mauro Castelli" ], "date": "2026/08/07", "category": "Data and representations", "origin": "Linked in seed", "priority": "Supporting", "depth": "Methods/results", "sensing": "Human motion / vision", "contribution": "Uses explicit kinematic retargeting of human demonstrations to improve VLA robustness.", "boundary": "Relevant to demonstration conversion; tactile supervision is not established by this source.", "url": "https://www.dma-journal.org/index.php/dema/article/view/180", "fulltext_url": "", "accessed": "2026-09-08" }, { "id": "P07", "alias": "OpenVLA", "title": "OpenVLA: An Open-Source Vision-Language-Action Model", "authors": [ "Kim, Moo Jin", "Pertsch, Karl", "Karamcheti, Siddharth", "Xiao, Ted", "Balakrishna, Ashwin", "Nair, Suraj", "Rafailov, Rafael", "Foster, Ethan", "Lam, Grace", "Sanketi, Pannag", "Vuong, Quan", "Kollar, Thomas", "Burchfiel, Benjamin", "Tedrake, Russ", "Sadigh, Dorsa", "Levine, Sergey", "Liang, Percy", "Finn, Chelsea" ], "date": "2024/06/13", "category": "Policy and fusion", "origin": "Linked in seed", "priority": "Supporting", "depth": "Abstract/project", "sensing": "Vision + language + actions", "contribution": "Open VLA training and adaptation reference.", "boundary": "A possible backbone, not a native pressure-array data generator.", "url": "https://arxiv.org/abs/2406.09246", "fulltext_url": "", "accessed": "2026-09-08" }, { "id": "P08", "alias": "G0.5", "title": "G0.5: One Autoregressive Stream for Robot Reasoning and Action", "authors": [ "Liu, Yicheng", "Dong, Zibin", "Ye, Baijun", "Yuan, Tianyuan", "Jiang, Tao", "Yang, Anqi", "Cao, Shicheng", "Liu, Haonan", "Sun, Yue", "Guo, Zihan", "Liu, Xiao", "Ke, Dong", "Pan, Changxun", "Wu, Chenru", "Cheng, Tailai", "Ren, Xiaoshu", "Zhang, Xinlei", "Cui, Jianning", "Zhao, Zijie", "Zhang, Haoyu", "Xu, Kaiming", "Yang, Haodong", "Zhang, Bowen", "Niu, Jiahui", "Zhu, Shaoting", "Zhang, Shiduo", "Zhao, Hang" ], "date": "2026/08/12", "category": "Policy and fusion", "origin": "Linked in seed", "priority": "Supporting", "depth": "Abstract/project", "sensing": "Vision / reasoning / actions", "contribution": "Combines robot reasoning and action in an autoregressive stream.", "boundary": "GalaxeaVLA is its companion repository, not an additional paper; tactile support must be separately established.", "url": "https://arxiv.org/abs/2608.11739", "fulltext_url": "", "accessed": "2026-09-08" }, { "id": "P09", "alias": "ASPIRE", "title": "ASPIRE: Agentic /Skills Discovery for Robotics", "authors": [ "Lu, Runyu", "Wu, Yubo", "Kou, Ethan", "Fu, Letian", "Xiao, Wenli", "Mandlekar, Ajay", "Xu, Yinzhen", "Shi, Guanya", "Goldberg, Ken", "Chen, Ang", "Chowdhury, Mosharaf", "Zhu, Yuke", "Fan, Linxi \"Jim\"", "Wang, Guanzhi" ], "date": "2026/06/30", "category": "Agents and generation", "origin": "Linked in seed", "priority": "Core", "depth": "Methods/results + code", "sensing": "Vision + executable robot programs", "contribution": "Discovers, repairs, and reuses executable robot skills with an agent.", "boundary": "Useful teacher-generation precedent; code-policy success is distinct from training a neural tactile policy.", "url": "https://arxiv.org/abs/2607.00272", "fulltext_url": "https://arxiv.org/html/2607.00272v1", "accessed": "2026-09-08" }, { "id": "P10", "alias": "ENPIRE", "title": "ENPIRE: Agentic Robot Policy Self-Improvement in the Real World", "authors": [ "Xiao, Wenli", "Xie, Jia", "Zhang, Tonghe", "Lin, Haotian", "Fu, Letian \"Max\"", "Xue, Haoru", "Lu, Jalen", "Yang, Yi", "Dai, Cunxi", "Wang, Zi", "Wu, Jimmy", "Wang, Guanzhi", "Sastry, S. Shankar", "Goldberg, Ken", "Fan, Linxi \"Jim\"", "Zhu, Yuke", "Shi, Guanya" ], "date": "2026/06/18", "category": "Agents and generation", "origin": "Linked in seed", "priority": "Core", "depth": "Methods/results + code", "sensing": "Vision and robot feedback", "contribution": "Improves robot code policies through real execution, verification, and resets.", "boundary": "Gripper torque feedback is not a spatial tactile array; repeated-attempt success must not be reported as single-attempt success.", "url": "https://arxiv.org/abs/2606.19980", "fulltext_url": "https://arxiv.org/html/2606.19980v1", "accessed": "2026-09-08" }, { "id": "P11", "alias": "SimFoundry", "title": "SimFoundry: Modular and Automated Scene Generation for Policy Learning and Evaluation", "authors": [ "Ranawaka, Nadun", "Wong, Josiah", "Pai, Wei-Lin", "Chu, Wei-Teng", "Dai, Tianyuan", "Moghani, Masoud", "Yin, Hang", "Jiang, Yunfan", "Durbano, Wesley", "Huynh, Brandon", "Fang, Yu", "Xu, Danfei", "Zhang, Ruohan", "Fei-Fei, Li", "Fan, Linxi", "Wen, Bowen", "Mandlekar, Ajay", "Zhu, Yuke" ], "date": "2026/06/26", "category": "Agents and generation", "origin": "Linked in seed", "priority": "Core", "depth": "Methods/results + code", "sensing": "Visual scene generation", "contribution": "Builds modular simulation scenes for policy learning and evaluation.", "boundary": "Scene assets help diversity; estimated physical properties require calibration and release scope is narrower than the full paper pipeline.", "url": "https://arxiv.org/abs/2606.28276", "fulltext_url": "https://arxiv.org/html/2606.28276v1", "accessed": "2026-09-08" }, { "id": "P12", "alias": "MolmoAct2", "title": "MolmoAct2: Action Reasoning Models for Real-world Deployment", "authors": [ "Fang, Haoquan", "Duan, Jiafei", "Clay, Donovan", "Wang, Sam", "Liu, Shuo", "Huang, Weikai", "Fan, Xiang", "Tsai, Wei-Chuan", "Chen, Shirui", "Wang, Yi Ru", "Xing, Shanli", "Cho, Jaemin", "Park, Jae Sung", "Eftekhar, Ainaz", "Sushko, Peter", "Farley, Karen", "Wadhwa, Angad", "Harrison, Cole", "Han, Winson", "Lee, Ying-Chun", "VanderBilt, Eli", "Hendrix, Rose", "Ellawela, Suveen", "Ngoo, Lucas", "Chai, Joyce", "Ren, Zhongzheng", "Farhadi, Ali", "Fox, Dieter", "Krishna, Ranjay" ], "date": "2026/05/04", "category": "Policy and fusion", "origin": "Linked in seed", "priority": "Supporting", "depth": "Abstract/project + code", "sensing": "Vision + reasoning + actions", "contribution": "Action reasoning model with real-robot deployment workflows.", "boundary": "Candidate student backbone; inspected release did not establish native spatial pressure input.", "url": "https://arxiv.org/abs/2605.02881", "fulltext_url": "", "accessed": "2026-09-08" }, { "id": "P13", "alias": "ACT / ALOHA", "title": "Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardware", "authors": [ "Zhao, Tony Z.", "Kumar, Vikash", "Levine, Sergey", "Finn, Chelsea" ], "date": "2023/04/23", "category": "Policy and fusion", "origin": "Named in seed", "priority": "Supporting", "depth": "Abstract/project", "sensing": "Vision + proprioception + actions", "contribution": "Introduces action chunking for fine-grained bimanual imitation learning with low-cost hardware.", "boundary": "A compact student baseline can reveal whether synthetic data helps before large VLA training.", "url": "https://arxiv.org/abs/2304.13705", "fulltext_url": "", "accessed": "2026-09-08" }, { "id": "P14", "alias": "FlexiTac", "title": "FlexiTac: A Low-Cost, Open-Source, Scalable Tactile Sensing Solution for Robotic Systems", "authors": [ "Huang, Binghao", "Li, Yunzhu" ], "date": "2026/04/30", "category": "Simulation and sensors", "origin": "Expanded search; sensor named in seed", "priority": "Core", "depth": "Abstract/project + code", "sensing": "Piezoresistive pressure arrays", "contribution": "Provides a low-cost scalable tactile sensing platform.", "boundary": "Keep separate from optical GelSight/DIGIT simulators; the IsaacSim example is not the calibrated VT-Refine pipeline.", "url": "https://arxiv.org/abs/2604.28156", "fulltext_url": "", "accessed": "2026-09-08" }, { "id": "P15", "alias": "UniVTAC", "title": "UniVTAC: A Unified Simulation Platform for Visuo-Tactile Manipulation Data Generation, Learning, and Benchmarking", "authors": [ "Chen, Baijun", "Wan, Weijie", "Chen, Tianxing", "Guo, Xianda", "Xu, Congsheng", "Qi, Yuanyang", "Zhang, Haojie", "Wu, Longyan", "Xu, Tianling", "Li, Zixuan", "Wu, Yizhe", "Li, Rui", "Yang, Xiaokang", "Luo, Ping", "Sui, Wei", "Mu, Yao" ], "date": "2026/02/10", "category": "Simulation and sensors", "origin": "Expanded search", "priority": "Core", "depth": "Methods/results", "sensing": "Optical tactile + vision + actions", "contribution": "Unifies visuo-tactile simulation, automated data generation, correction, learning, and benchmarking.", "boundary": "Direct prior art for synthetic tactile data and contact checks; adaptation to FlexiTac requires a different sensor model.", "url": "https://arxiv.org/abs/2602.10093", "fulltext_url": "https://arxiv.org/html/2602.10093v1", "accessed": "2026-09-08" }, { "id": "P16", "alias": "TACTO", "title": "TACTO: A Fast, Flexible, and Open-source Simulator for High-Resolution Vision-based Tactile Sensors", "authors": [ "Wang, Shaoxiong", "Lambeta, Mike", "Chou, Po-Wei", "Calandra, Roberto" ], "date": "2020/12/15", "category": "Simulation and sensors", "origin": "Expanded search", "priority": "Supporting", "depth": "Abstract/project", "sensing": "Optical tactile rendering", "contribution": "Provides a fast renderer and interfaces for simulating vision-based tactile sensors.", "boundary": "Useful rendering baseline; visual similarity does not by itself establish accurate contact dynamics.", "url": "https://arxiv.org/abs/2012.08456", "fulltext_url": "", "accessed": "2026-09-08" }, { "id": "P17", "alias": "Taxim", "title": "Taxim: An Example-based Simulation Model for GelSight Tactile Sensors", "authors": [ "Si, Zilin", "Yuan, Wenzhen" ], "date": "2021/09/09", "category": "Simulation and sensors", "origin": "Expanded search", "priority": "Supporting", "depth": "Abstract/project", "sensing": "GelSight optical response and markers", "contribution": "Calibrates optical rendering and models marker displacement using an example-based simulator.", "boundary": "Sensor rendering component rather than a complete environment and teacher generator.", "url": "https://arxiv.org/abs/2109.04027", "fulltext_url": "", "accessed": "2026-09-08" }, { "id": "P18", "alias": "Tacchi", "title": "Tacchi: A Pluggable and Low Computational Cost Elastomer Deformation Simulator for Optical Tactile Sensors", "authors": [ "Chen, Zixi", "Zhang, Shixin", "Luo, Shan", "Sun, Fuchun", "Fang, Bin" ], "date": "2023/01/19", "category": "Simulation and sensors", "origin": "Expanded search", "priority": "Supporting", "depth": "Abstract/project", "sensing": "Optical elastomer deformation", "contribution": "Uses particles and Taichi for a pluggable tactile deformation simulator.", "boundary": "Compare deformation cost and fidelity; full robot integration and sensor calibration remain necessary.", "url": "https://arxiv.org/abs/2301.08343", "fulltext_url": "", "accessed": "2026-09-08" }, { "id": "P19", "alias": "TacSL", "title": "TacSL: A Library for Visuotactile Sensor Simulation and Learning", "authors": [ "Akinola, Iretiayo", "Xu, Jie", "Carius, Jan", "Fox, Dieter", "Narang, Yashraj" ], "date": "2024/08/12", "category": "Simulation and sensors", "origin": "Expanded search", "priority": "Core", "depth": "Abstract/project", "sensing": "Visuotactile images and contact-force distributions", "contribution": "GPU-parallel tactile simulation and policy learning with sim-to-real transfer.", "boundary": "Central to the VT-Refine implementation lineage; Isaac Gym and Isaac Sim interfaces are not interchangeable.", "url": "https://arxiv.org/abs/2408.06506", "fulltext_url": "", "accessed": "2026-09-08" }, { "id": "P20", "alias": "TacEx", "title": "TacEx: GelSight Tactile Simulation in Isaac Sim -- Combining Soft-Body and Visuotactile Simulators", "authors": [ "Nguyen, Duc Huy", "Schneider, Tim", "Duret, Guillaume", "Kshirsagar, Alap", "Belousov, Boris", "Peters, Jan" ], "date": "2024/11/07", "category": "Simulation and sensors", "origin": "Expanded search", "priority": "Core", "depth": "Abstract/project", "sensing": "Optical tactile, soft contacts", "contribution": "Combines deformable contact simulation with tactile rendering in Isaac Sim.", "boundary": "A platform candidate for optical sensors; validate throughput and supported sensor configuration in the intended workload.", "url": "https://arxiv.org/abs/2411.04776", "fulltext_url": "", "accessed": "2026-09-08" }, { "id": "P21", "alias": "Taccel", "title": "Taccel: Scaling Up Vision-based Tactile Robotics via High-performance GPU Simulation", "authors": [ "Li, Yuyang", "Du, Wenxin", "Yu, Chang", "Li, Puhao", "Zhao, Zihang", "Liu, Tengyu", "Jiang, Chenfanfu", "Zhu, Yixin", "Huang, Siyuan" ], "date": "2025/04/17", "category": "Simulation and sensors", "origin": "Expanded search", "priority": "Core", "depth": "Methods/results", "sensing": "Vision-based tactile with soft gel dynamics", "contribution": "Uses GPU IPC and affine-body dynamics for parallel robot and tactile simulation.", "boundary": "Strong optical physics baseline; published scaling conditions should not be assumed for a different GPU or task.", "url": "https://arxiv.org/abs/2504.12908", "fulltext_url": "https://arxiv.org/html/2504.12908v1", "accessed": "2026-09-08" }, { "id": "P22", "alias": "DiffTactile", "title": "DIFFTACTILE: A Physics-based Differentiable Tactile Simulator for Contact-rich Robotic Manipulation", "authors": [ "Si, Zilin", "Zhang, Gu", "Ben, Qingwei", "Romero, Branden", "Xian, Zhou", "Liu, Chao", "Gan, Chuang" ], "date": "2024/03/13", "category": "Simulation and sensors", "origin": "Expanded search", "priority": "Supporting", "depth": "Abstract/project", "sensing": "Differentiable contact and tactile signals", "contribution": "Differentiable physics for contact-rich manipulation with deformable sensors and objects.", "boundary": "Relevant to calibration and control optimization; scalability for large agent-generated corpora is unestablished here.", "url": "https://arxiv.org/abs/2403.08716", "fulltext_url": "", "accessed": "2026-09-08" }, { "id": "P23", "alias": "ManiFeel", "title": "ManiFeel: Benchmarking and Understanding Visuotactile Manipulation Policy Learning", "authors": [ "Luu, Quan Khanh", "Zhou, Pokuang", "Xu, Zhengtong", "Zhang, Zhiyuan", "Qiu, Qiang", "She, Yu" ], "date": "2025/05/24", "category": "Simulation and sensors", "origin": "Expanded search", "priority": "Core", "depth": "Methods/results", "sensing": "Multiple visuotactile representations", "contribution": "Benchmarks supervised visuotactile policies across contact-rich and visually challenging tasks.", "boundary": "Real-world policies were trained on real datasets to compare modality trends; these experiments are not direct simulation-to-real policy transfer.", "url": "https://arxiv.org/abs/2505.18472", "fulltext_url": "https://arxiv.org/html/2505.18472v2", "accessed": "2026-09-08" }, { "id": "P24", "alias": "IsaacIPC", "title": "IsaacIPC: Coupling High-Fidelity Simulation and Realistic Rendering for Contact-Rich Robotic Systems", "authors": [ "Liang, Qixin", "Han, Zhongqing" ], "date": "2026/05/23", "category": "Simulation and sensors", "origin": "Expanded search", "priority": "Supporting", "depth": "Abstract/project", "sensing": "Contact physics + rendering", "contribution": "Couples GPU IPC with Isaac Sim/Lab and introduces a tactile-surface contact potential to improve contact-pressure distributions.", "boundary": "Candidate for difficult contacts and deformation, not proof that arbitrary generated tasks transfer to real hardware.", "url": "https://arxiv.org/abs/2605.24339", "fulltext_url": "", "accessed": "2026-09-08" }, { "id": "P25", "alias": "RoboPack", "title": "RoboPack: Learning Tactile-Informed Dynamics Models for Dense Packing", "authors": [ "Ai, Bo", "Tian, Stephen", "Shi, Haochen", "Wang, Yixuan", "Tan, Cheston", "Li, Yunzhu", "Wu, Jiajun" ], "date": "2024/07/01", "category": "World models", "origin": "Expanded search", "priority": "Core", "depth": "Methods/results", "sensing": "Soft-Bubble tactile + visual geometry", "contribution": "Learns tactile-informed dynamics and latent physical information for dense packing and planning.", "boundary": "Yunzhu Li research lineage; structured latent physics is an alternative to generating tactile video frames.", "url": "https://arxiv.org/abs/2407.01418", "fulltext_url": "https://arxiv.org/html/2407.01418v1", "accessed": "2026-09-08" }, { "id": "P26", "alias": "VT-WM", "title": "Visuo-Tactile World Models", "authors": [ "Higuera, Carolina", "Arnaud, Sergio", "Boots, Byron", "Mukadam, Mustafa", "Hogan, Francois Robert", "Meier, Franziska" ], "date": "2026/02/05", "category": "World models", "origin": "Expanded search", "priority": "Core", "depth": "Methods/results", "sensing": "DIGIT 360 tactile images + vision", "contribution": "Learns multi-task visuo-tactile predictive models and uses imagined rollouts for planning.", "boundary": "Planning uses the same scene and objects; generalization to new physical properties remains open. CEM planning is computationally costly.", "url": "https://arxiv.org/abs/2602.06001", "fulltext_url": "https://arxiv.org/html/2602.06001v1", "accessed": "2026-09-08" }, { "id": "P27", "alias": "ViTacWorld", "title": "ViTacWorld: Scaling Visuo-Tactile World Models for Contact-Rich Robot Manipulation", "authors": [ "Huang, Yunao", "Sang, Shiyu", "Lu, Haotao", "Ni, Suting", "Wu, Shijie", "Guo, Ziyang", "Shi, Ye", "Wang, Jingya" ], "date": "2026/07/24", "category": "World models", "origin": "Expanded search", "priority": "Core", "depth": "Methods/results", "sensing": "Optical tactile + vision + actions", "contribution": "Uses real and simulated interactions to learn action-conditioned visuo-tactile rollouts for policy augmentation and evaluation.", "boundary": "Very close to the proposed data-generation goal; real-rollout adaptation and generated-data selection remain material assumptions.", "url": "https://arxiv.org/abs/2607.22530", "fulltext_url": "https://arxiv.org/html/2607.22530v1", "accessed": "2026-09-08" }, { "id": "P28", "alias": "Dream-Tac", "title": "Dream-Tac: A Unified Tactile World Action Model for Contact-Rich Robot Manipulation", "authors": [ "Lou, Yunfan", "Ye, Yifan", "Fu, Yankai", "Cen, Jun", "Chi, Xiaowei", "Lyu, Yaoxu", "Jia, Peidong", "Han, Sirui", "Lu, Zhihe", "Zhang, Shanghang" ], "date": "2026/06/07", "category": "World models", "origin": "Expanded search", "priority": "Core", "depth": "Methods/results", "sensing": "Optical tactile + vision + actions", "contribution": "Unifies future visual/tactile prediction with action modeling.", "boundary": "Predicting touch inside a world-action model is established prior art; compare downstream transfer and failure recovery.", "url": "https://arxiv.org/abs/2606.08737", "fulltext_url": "https://arxiv.org/html/2606.08737v1", "accessed": "2026-09-08" }, { "id": "P29", "alias": "ContactWorld", "title": "ContactWorld: What Matters in Vision-Tactile World Models for Contact-Rich Manipulation", "authors": [ "Zhang, Zhiyuan", "Zhou, Pokuang", "Zhang, Kaidi", "Desai, Adeesh", "Amosa, Temitope", "Soleymanzadeh, Davood", "Lei, Jiuzhou", "Zheng, Minghui", "She, Yu" ], "date": "2026/06/11", "category": "World models", "origin": "Expanded search", "priority": "Core", "depth": "Methods/results", "sensing": "Vision / point clouds and tactile fields", "contribution": "Studies representation choices for contact-rich visuo-tactile world modeling.", "boundary": "Use to separate benefits of data generation from benefits of representation design.", "url": "https://arxiv.org/abs/2606.13877", "fulltext_url": "https://arxiv.org/html/2606.13877v1", "accessed": "2026-09-08" }, { "id": "P30", "alias": "OmniVTA / OmniViTac", "title": "OmniVTA: Visuo-Tactile World Modeling for Contact-Rich Robotic Manipulation", "authors": [ "Zheng, Yuhang", "Gu, Songen", "Zheng, Yupeng", "Li, Weize", "Zang, Yujie", "Tian, Shuai", "Li, Xiang", "Hao, Ce", "Gao, Chen", "Liu, Si", "Li, Haoran", "Chen, Yilun", "Yan, Shuicheng", "Ding, Wenchao" ], "date": "2026/03/19", "category": "World models", "origin": "Expanded search", "priority": "Core", "depth": "Methods/results", "sensing": "Optical tactile / displacement fields + vision + actions", "contribution": "Combines a large visuo-tactile-action dataset, predictive contact modeling, fusion, and fast corrective control.", "boundary": "OmniViTac is the dataset within this paper, not another paper; real data scale and embodiment need separate accounting.", "url": "https://arxiv.org/abs/2603.19201", "fulltext_url": "https://arxiv.org/html/2603.19201v3", "accessed": "2026-09-08" }, { "id": "P31", "alias": "Interactive World Simulator", "title": "Interactive World Simulator for Robot Policy Training and Evaluation", "authors": [ "Wang, Yixuan", "Syed, Rhythm", "Wu, Fangyu", "Zhang, Mengchao", "Onol, Aykut", "Barreiros, Jose", "Nayyeri, Hooshang", "Dear, Tony", "Zhang, Huan", "Li, Yunzhu" ], "date": "2026/03/09", "category": "World models", "origin": "Expanded search", "priority": "Core", "depth": "Abstract/project", "sensing": "Action-conditioned visual prediction", "contribution": "Yunzhu Li and collaborators study learned simulation for policy training and evaluation.", "boundary": "Relevant visual world-model baseline; tactile sensing is not established by the reviewed project material.", "url": "https://arxiv.org/abs/2603.08546", "fulltext_url": "", "accessed": "2026-09-08" }, { "id": "P32", "alias": "Tactile-VLA", "title": "Tactile-VLA: Unlocking Vision-Language-Action Model's Physical Knowledge for Tactile Generalization", "authors": [ "Huang, Jialei", "Wang, Shuo", "Lin, Fanqi", "Hu, Yihang", "Wen, Chuan", "Gao, Yang" ], "date": "2025/07/12", "category": "Policy and fusion", "origin": "Expanded search", "priority": "Core", "depth": "Methods/results", "sensing": "Normal/shear tactile feedback + vision + language", "contribution": "Connects VLA physical knowledge and tactile reasoning with hybrid position-force control.", "boundary": "Demonstrations include tactile-guided forces and failures; FlexiTac normal pressure does not directly reproduce the reported normal/shear sensing.", "url": "https://arxiv.org/abs/2507.09160", "fulltext_url": "https://arxiv.org/html/2507.09160v1", "accessed": "2026-09-08" }, { "id": "P33", "alias": "TacFiLM", "title": "Tactile Modality Fusion for Vision-Language-Action Models", "authors": [ "Morissette, Charlotte", "Abyaneh, Amin", "Chang, Wei-Di", "Houssaini, Anas", "Meger, David", "Lin, Hsiu-Chin", "Tremblay, Jonathan", "Dudek, Gregory" ], "date": "2026/03/15", "category": "Policy and fusion", "origin": "Expanded search", "priority": "Core", "depth": "Methods/results", "sensing": "Pretrained optical tactile features + vision", "contribution": "Modulates intermediate visual features with tactile representations for lightweight VLA adaptation.", "boundary": "Experiments use DIGIT and OpenVLA-OFT with LoRA; insertion starts with the object already held. Pressure-array compatibility remains untested.", "url": "https://arxiv.org/abs/2603.14604", "fulltext_url": "https://arxiv.org/html/2603.14604v1", "accessed": "2026-09-08" }, { "id": "P34", "alias": "ForceVLA2", "title": "ForceVLA2: Unleashing Hybrid Force-Position Control with Force Awareness for Contact-Rich Manipulation", "authors": [ "Li, Yang", "Zhaxizhuoma", "Jiang, Hongru", "Xia, Junjie", "Zhang, Hongquan", "Du, Jinda", "Zhou, Yunsong", "Zeng, Jia", "Hao, Ce", "Ren, Jieji", "Yu, Qiaojun", "Lu, Cewu", "Qiao, Yu", "Pang, Jiangmiao" ], "date": "2026/03/16", "category": "Policy and fusion", "origin": "Expanded search", "priority": "Supporting", "depth": "Abstract/project", "sensing": "Force awareness + hybrid position-force actions", "contribution": "Builds force-aware VLA control for contact-rich tasks.", "boundary": "Include a force-feedback control baseline; force/torque sensing is distinct from dense tactile arrays.", "url": "https://arxiv.org/abs/2603.15169", "fulltext_url": "", "accessed": "2026-09-08" }, { "id": "P35", "alias": "OmniVTLA", "title": "OmniVTLA: Vision-Tactile-Language-Action Models with Semantic-Aligned Tactile Sensing", "authors": [ "Cheng, Zhengxue", "Zhang, Yiqian", "Tang, Anni", "Wang, Keyu", "Zhang, Wenkang", "Li, Haoyu", "Zhang, Hengdi", "Song, Li" ], "date": "2025/08/12", "category": "Policy and fusion", "origin": "Expanded search", "priority": "Supporting", "depth": "Abstract/project", "sensing": "Optical and force-based tactile", "contribution": "Uses semantic-aligned tactile representations and introduces the ObjTac dataset.", "boundary": "Cross-sensor fusion reference; object/property samples are not automatically full action-labeled trajectories.", "url": "https://arxiv.org/abs/2508.08706", "fulltext_url": "", "accessed": "2026-09-08" }, { "id": "P36", "alias": "TouchGuide", "title": "TouchGuide: Inference-Time Steering of Visuomotor Policies via Touch Guidance", "authors": [ "Zhang, Zhemeng", "Ma, Jiahua", "Yang, Xincheng", "Wen, Xin", "Zhang, Yuzhi", "Li, Boyan", "Qin, Yiran", "Liu, Jin", "Zhao, Can", "Kang, Li", "Hong, Haoqin", "Yin, Zhenfei", "Torr, Philip", "Su, Hao", "Zhang, Ruimao", "Ma, Daolin" ], "date": "2026/01/28", "category": "Policy and fusion", "origin": "Expanded search", "priority": "Core", "depth": "Abstract/project", "sensing": "Tactile-guided action refinement", "contribution": "Uses a contact model to guide diffusion/flow policy actions at inference time.", "boundary": "Tests whether synthetic contact data helps a small corrective model without retraining an entire VLA.", "url": "https://arxiv.org/abs/2601.20239", "fulltext_url": "", "accessed": "2026-09-08" }, { "id": "P37", "alias": "Reactive Diffusion Policy", "title": "Reactive Diffusion Policy: Slow-Fast Visual-Tactile Policy Learning for Contact-Rich Manipulation", "authors": [ "Xue, Han", "Ren, Jieji", "Chen, Wendi", "Zhang, Gu", "Fang, Yuan", "Gu, Guoying", "Xu, Huazhe", "Lu, Cewu" ], "date": "2025/03/04", "category": "Policy and fusion", "origin": "Expanded search", "priority": "Core", "depth": "Abstract/project", "sensing": "Visual planning + fast tactile/force feedback", "contribution": "Combines slower action-chunk planning with fast contact-responsive control.", "boundary": "A temporal control baseline; action chunking alone can miss contact changes during execution.", "url": "https://arxiv.org/abs/2503.02881", "fulltext_url": "", "accessed": "2026-09-08" }, { "id": "P38", "alias": "FACTR", "title": "FACTR: Force-Attending Curriculum Training for Contact-Rich Policy Learning", "authors": [ "Liu, Jason Jingzhou", "Li, Yulong", "Shaw, Kenneth", "Tao, Tony", "Salakhutdinov, Ruslan", "Pathak, Deepak" ], "date": "2025/02/24", "category": "Policy and fusion", "origin": "Expanded search", "priority": "Core", "depth": "Abstract/project", "sensing": "Robot force feedback + vision", "contribution": "Uses a visual corruption curriculum to encourage policy attention to force.", "boundary": "Critical comparison for a proposed modality curriculum; this work does not generate new environments.", "url": "https://arxiv.org/abs/2502.17432", "fulltext_url": "", "accessed": "2026-09-08" }, { "id": "P39", "alias": "VLA-Touch", "title": "VLA-Touch: Enhancing Vision-Language-Action Models with Dual-Level Tactile Feedback", "authors": [ "Bi, Jianxin", "Ma, Kevin Yuchen", "Hao, Ce", "Shou, Mike Zheng", "Soh, Harold" ], "date": "2025/07/23", "category": "Policy and fusion", "origin": "Expanded search", "priority": "Core", "depth": "Abstract/project", "sensing": "Tactile semantics + low-level touch control", "contribution": "Adds high-level tactile feedback and diffusion-based action refinement around a frozen VLA.", "boundary": "A modular integration baseline; distinct from Tactile-VLA despite similar names.", "url": "https://arxiv.org/abs/2507.17294", "fulltext_url": "", "accessed": "2026-09-08" }, { "id": "P40", "alias": "FreeTacMan", "title": "FreeTacMan: Robot-free Visuo-Tactile Data Collection System for Contact-rich Manipulation", "authors": [ "Wu, Longyan", "Yu, Checheng", "Ren, Jieji", "Chen, Li", "Jiang, Yufei", "Huang, Ran", "Gu, Guoying", "Li, Hongyang" ], "date": "2025/06/02", "category": "Data and representations", "origin": "Expanded search", "priority": "Core", "depth": "Abstract/project", "sensing": "Robot-free optical visuotactile demonstrations", "contribution": "Collects contact-rich demonstrations with a handheld gripper and supports tactile pretraining.", "boundary": "Real-data calibration and co-training source; robot-free collection does not mean synthetic data.", "url": "https://arxiv.org/abs/2506.01941", "fulltext_url": "", "accessed": "2026-09-08" }, { "id": "P41", "alias": "exUMI", "title": "exUMI: Extensible Robot Teaching System with Action-aware Task-agnostic Tactile Representation", "authors": [ "Xu, Yue", "Wei, Litao", "An, Pengyu", "Zhang, Qingyu", "Li, Yong-Lu" ], "date": "2025/09/18", "category": "Data and representations", "origin": "Expanded search", "priority": "Core", "depth": "Abstract/project", "sensing": "9DTact + pose / gripper motion", "contribution": "Extends UMI and pretrains tactile representations through action-aware temporal prediction.", "boundary": "Direct alternative for obtaining action-conditioned tactile data; check action and coordinate conversion across embodiments.", "url": "https://arxiv.org/abs/2509.14688", "fulltext_url": "", "accessed": "2026-09-08" }, { "id": "P42", "alias": "TacUMI (Cheng et al.)", "title": "TacUMI: A Multi-Modal Universal Manipulation Interface for Contact-Rich Tasks", "authors": [ "Cheng, Tailai", "Chen, Kejia", "Chen, Lingyun", "Zhang, Liding", "Zhang, Yue", "Ling, Yao", "Hamad, Mahdi", "Bing, Zhenshan", "Wu, Fan", "Sharma, Karan", "Knoll, Alois" ], "date": "2026/01/21", "category": "Data and representations", "origin": "Expanded search", "priority": "Supporting", "depth": "Abstract/project", "sensing": "ViTac + force/torque + pose", "contribution": "Synchronizes multimodal handheld demonstrations and segments contact-rich sequences.", "boundary": "Name collision: this paper differs from the TacUMI devices described by TouchGuide and OmniVTA.", "url": "https://arxiv.org/abs/2601.14550", "fulltext_url": "", "accessed": "2026-09-08" }, { "id": "P43", "alias": "ViTaMIn", "title": "ViTaMIn: Learning Contact-Rich Tasks Through Robot-Free Visuo-Tactile Manipulation Interface", "authors": [ "Liu, Fangchen", "Li, Chuanyu", "Qin, Yihua", "Xu, Jing", "Abbeel, Pieter", "Chen, Rui" ], "date": "2025/04/08", "category": "Data and representations", "origin": "Expanded search", "priority": "Supporting", "depth": "Abstract/project", "sensing": "Compliant Fin Ray gripper + visuotactile sensing", "contribution": "Robot-free demonstration collection and multimodal representation learning.", "boundary": "Its compliant gripper changes contact mechanics; demonstration transfer needs more than matching camera views.", "url": "https://arxiv.org/abs/2504.06156", "fulltext_url": "", "accessed": "2026-09-08" }, { "id": "P44", "alias": "Sparsh", "title": "Sparsh: Self-supervised touch representations for vision-based tactile sensing", "authors": [ "Higuera, Carolina", "Sharma, Akash", "Bodduluri, Chaithanya Krishna", "Fan, Taosha", "Lancaster, Patrick", "Kalakrishnan, Mrinal", "Kaess, Michael", "Boots, Byron", "Lambeta, Mike", "Wu, Tingfan", "Mukadam, Mustafa" ], "date": "2024/10/31", "category": "Data and representations", "origin": "Expanded search", "priority": "Core", "depth": "Abstract/project", "sensing": "Optical tactile images", "contribution": "Self-supervised tactile encoders with a multi-task TacBench evaluation suite.", "boundary": "Useful optical encoder initialization; frame/sample counts should not be presented as robot trajectories.", "url": "https://arxiv.org/abs/2410.24090", "fulltext_url": "", "accessed": "2026-09-08" }, { "id": "P45", "alias": "Sparsh-X", "title": "Tactile Beyond Pixels: Multisensory Touch Representations for Robot Manipulation", "authors": [ "Higuera, Carolina", "Sharma, Akash", "Fan, Taosha", "Bodduluri, Chaithanya Krishna", "Boots, Byron", "Kaess, Michael", "Lambeta, Mike", "Wu, Tingfan", "Liu, Zixi", "Hogan, Francois Robert", "Mukadam, Mustafa" ], "date": "2025/06/17", "category": "Data and representations", "origin": "Expanded search", "priority": "Supporting", "depth": "Abstract/project", "sensing": "DIGIT 360 image, audio, motion, pressure", "contribution": "Learns multisensory tactile representations and studies manipulation transfer.", "boundary": "Requires a richer sensor suite than FlexiTac; do not assume all pretraining inputs exist on the target robot.", "url": "https://arxiv.org/abs/2506.14754", "fulltext_url": "", "accessed": "2026-09-08" }, { "id": "P46", "alias": "UniTouch", "title": "Binding Touch to Everything: Learning Unified Multimodal Tactile Representations", "authors": [ "Yang, Fengyu", "Feng, Chao", "Chen, Ziyang", "Park, Hyoungseob", "Wang, Daniel", "Dou, Yiming", "Zeng, Ziyao", "Chen, Xien", "Gangopadhyay, Rit", "Owens, Andrew", "Wong, Alex" ], "date": "2024/01/31", "category": "Data and representations", "origin": "Expanded search", "priority": "Supporting", "depth": "Abstract/project", "sensing": "Optical touch aligned with multimodal embeddings", "contribution": "Aligns heterogeneous tactile sensors to visual representations connected with language and sound.", "boundary": "Perception and semantic alignment reference, not an autonomous action-data generator.", "url": "https://arxiv.org/abs/2401.18084", "fulltext_url": "", "accessed": "2026-09-08" }, { "id": "P47", "alias": "AnyTouch 2", "title": "AnyTouch 2: General Optical Tactile Representation Learning For Dynamic Tactile Perception", "authors": [ "Feng, Ruoxuan", "Zhou, Yuxuan", "Mei, Siyu", "Zhou, Dongzhan", "Wang, Pengwei", "Cui, Shaowei", "Fang, Bin", "Yao, Guocai", "Hu, Di" ], "date": "2026/02/10", "category": "Data and representations", "origin": "Expanded search", "priority": "Supporting", "depth": "Abstract/project", "sensing": "Dynamic optical tactile and force data", "contribution": "Studies general dynamic tactile representations and introduces ToucHD.", "boundary": "Useful dynamics-oriented encoder benchmark; verify task-specific data access and sensor coverage.", "url": "https://arxiv.org/abs/2602.09617", "fulltext_url": "", "accessed": "2026-09-08" }, { "id": "P48", "alias": "Octopi", "title": "Octopi: Object Property Reasoning with Large Tactile-Language Models", "authors": [ "Yu, Samson", "Lin, Kelvin", "Xiao, Anxing", "Duan, Jiafei", "Soh, Harold" ], "date": "2024/05/05", "category": "Data and representations", "origin": "Expanded search", "priority": "Supporting", "depth": "Abstract/project", "sensing": "Tactile video + language / object properties", "contribution": "Reasons about object properties with tactile-language models and a tactile video dataset.", "boundary": "Useful for semantic grounding; property annotations do not supply robot action trajectories.", "url": "https://arxiv.org/abs/2405.02794", "fulltext_url": "", "accessed": "2026-09-08" }, { "id": "P49", "alias": "Touch and Go", "title": "Touch and Go: Learning from Human-Collected Vision and Touch", "authors": [ "Yang, Fengyu", "Ma, Chenyang", "Zhang, Jiacheng", "Zhu, Jing", "Yuan, Wenzhen", "Owens, Andrew" ], "date": "2022/11/22", "category": "Data and representations", "origin": "Expanded search", "priority": "Supporting", "depth": "Abstract/project", "sensing": "Human-collected visual and tactile observations", "contribution": "Provides natural paired vision-touch data for multisensory learning.", "boundary": "Perception pretraining source; not a robot demonstration dataset with calibrated control actions.", "url": "https://arxiv.org/abs/2211.12498", "fulltext_url": "", "accessed": "2026-09-08" }, { "id": "P50", "alias": "RoboGen", "title": "RoboGen: Towards Unleashing Infinite Data for Automated Robot Learning via Generative Simulation", "authors": [ "Wang, Yufei", "Xian, Zhou", "Chen, Feng", "Wang, Tsun-Hsuan", "Wang, Yian", "Fragkiadaki, Katerina", "Erickson, Zackory", "Held, David", "Gan, Chuang" ], "date": "2023/11/02", "category": "Agents and generation", "origin": "Expanded search", "priority": "Core", "depth": "Methods/results", "sensing": "Task, scene and supervision programs", "contribution": "Automates robotic skill learning through generative simulation.", "boundary": "Direct conceptual prior art; adding a tactile channel alone needs a stronger empirical contribution.", "url": "https://arxiv.org/abs/2311.01455", "fulltext_url": "https://proceedings.mlr.press/v235/wang24cc.html", "accessed": "2026-09-08" }, { "id": "P51", "alias": "GenSim", "title": "GenSim: Generating Robotic Simulation Tasks via Large Language Models", "authors": [ "Wang, Lirui", "Ling, Yiyang", "Yuan, Zhecheng", "Shridhar, Mohit", "Bao, Chen", "Qin, Yuzhe", "Wang, Bailin", "Xu, Huazhe", "Wang, Xiaolong" ], "date": "2023/10/02", "category": "Agents and generation", "origin": "Expanded search", "priority": "Core", "depth": "Abstract/project", "sensing": "Executable simulation tasks and demonstrations", "contribution": "Uses coding LLMs to generate task curricula and expert demonstrations.", "boundary": "Task-level diversity is already an explicit objective; compare contact-conditioned generation beyond generic task variation.", "url": "https://arxiv.org/abs/2310.01361", "fulltext_url": "", "accessed": "2026-09-08" }, { "id": "P52", "alias": "GenSim2", "title": "GenSim2: Scaling Robot Data Generation with Multi-modal and Reasoning LLMs", "authors": [ "Hua, Pu", "Liu, Minghuan", "Macaluso, Annabella", "Lin, Yunfeng", "Zhang, Weinan", "Xu, Huazhe", "Wang, Lirui" ], "date": "2024/10/04", "category": "Agents and generation", "origin": "Expanded search", "priority": "Core", "depth": "Abstract/project", "sensing": "Articulated simulation + planning/RL teachers", "contribution": "Generates more realistic and articulated tasks with multimodal reasoning and policy training.", "boundary": "Strong environment-and-teacher baseline; tactile simulation and calibration remain additional work.", "url": "https://arxiv.org/abs/2410.03645", "fulltext_url": "", "accessed": "2026-09-08" }, { "id": "P53", "alias": "RoboTwin 2.0", "title": "RoboTwin 2.0: A Scalable Data Generator and Benchmark with Strong Domain Randomization for Robust Bimanual Robotic Manipulation", "authors": [ "Chen, Tianxing", "Chen, Zanxin", "Chen, Baijun", "Cai, Zijian", "Liu, Yibin", "Li, Zixuan", "Liang, Qiwei", "Lin, Xianliang", "Ge, Yiheng", "Gu, Zhenyu", "Deng, Weiliang", "Guo, Yubin", "Nian, Tian", "Xie, Xuanbing", "Chen, Qiangyu", "Su, Kailun", "Xu, Tianling", "Liu, Guodong", "Hu, Mengkang", "Gao, Huan-ang", "Wang, Kaixuan", "Liang, Zhixuan", "Qin, Yusen", "Yang, Xiaokang", "Luo, Ping", "Mu, Yao" ], "date": "2025/06/22", "category": "Agents and generation", "origin": "Expanded search", "priority": "Core", "depth": "Abstract/project", "sensing": "Bimanual simulation and execution code", "contribution": "Combines generated expert code with simulation feedback and structured domain randomization.", "boundary": "Mandatory baseline for code-generated data; existing visual data scaling is not evidence of tactile necessity.", "url": "https://arxiv.org/abs/2506.18088", "fulltext_url": "", "accessed": "2026-09-08" }, { "id": "P54", "alias": "MimicGen", "title": "MimicGen: A Data Generation System for Scalable Robot Learning using Human Demonstrations", "authors": [ "Mandlekar, Ajay", "Nasiriany, Soroush", "Wen, Bowen", "Akinola, Iretiayo", "Narang, Yashraj", "Fan, Linxi", "Zhu, Yuke", "Fox, Dieter" ], "date": "2023/10/26", "category": "Agents and generation", "origin": "Expanded search", "priority": "Core", "depth": "Abstract/project", "sensing": "Transformed robot demonstrations", "contribution": "Scales human demonstrations through object-centric transformations in simulation.", "boundary": "Use as a fixed data-generation baseline; transformed trajectories still need contact validity checks.", "url": "https://arxiv.org/abs/2310.17596", "fulltext_url": "", "accessed": "2026-09-08" }, { "id": "P55", "alias": "ChronoAgentic (formerly Coding Agent Is Good As World Simulator)", "title": "ChronoAgentic: A Code-based Multi-Agent World Simulator for Physically Grounded Simulation Construction", "authors": [ "Wang, Hongyu", "Wang, Jingquan", "Anilkumar, Ashvin", "Zou, Bocheng", "Serban, Radu", "Negrut, Dan" ], "date": "2026/05/14", "category": "Agents and generation", "origin": "Expanded search", "priority": "Core", "depth": "Methods/results", "sensing": "Executable PyChrono scenes + visual and physics review", "contribution": "Plans, writes, executes and repairs physics simulation code using multiple reviewers.", "boundary": "Latest v3 evaluates 80 physics demos; video judging has shared-model and clip-length confounds and does not validate calibrated tactile policy transfer.", "url": "https://arxiv.org/abs/2605.14398", "fulltext_url": "https://arxiv.org/html/2605.14398v3", "accessed": "2026-09-08" }, { "id": "P56", "alias": "Code World Model", "title": "Code World Model: Coding Agent as World Brain", "authors": [ "Chen, Yiwen", "Lin, Guosheng", "Zhang, Chi" ], "date": "2026/08/26", "category": "Agents and generation", "origin": "Expanded search", "priority": "Supporting", "depth": "Methods/results", "sensing": "Persistent code state + proxy/video rendering", "contribution": "Separates executable world-state evolution from learned visual generation.", "boundary": "Architectural reference; generated visual observations are not calibrated tactile labels or contact-force validation.", "url": "https://arxiv.org/abs/2608.25927", "fulltext_url": "https://arxiv.org/html/2608.25927v1", "accessed": "2026-09-08" } ]