{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,3,16]],"date-time":"2026-03-16T07:12:55Z","timestamp":1773645175465,"version":"3.50.1"},"reference-count":24,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2026,3,16]],"date-time":"2026-03-16T00:00:00Z","timestamp":1773619200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Bioinform."],"abstract":"<jats:p>Spatial reasoning is essential for solving complex tasks in dynamic and high-dimensional environments. However, current training models for spatial tasks are computationally demanding and heavily reliant on human input. To address this gap, we present Snake-ML, a web-based simulation tool and proof-of-concept framework designed to demonstrate client-side training of spatial reasoning tasks. Snake-ML serves as an efficient and intuitive test bed for developing spatial navigation strategies in browser-based environments. We chose the snake game as our test bed because it is well suited for demonstrating spatial reasoning in low-dimensional visual spaces while remaining relevant to higher-dimensional tasks, compared to alternative methods. Through quantitative analysis, on the edge alone, Snake-ML achieves a 4.58\u00d7 speedup in model inference. Additionally, we developed a direct TensorFlow.js GPU pipeline that achieves up to a 32\u00d7 speedup in training time without any CPU\/GPU synchronization. This pipeline has the potential to improve many edge-based AI visualization projects. Snake-ML shows potential for adaptability to complex spatial tasks, such as autonomous systems, robotics, and AI-driven environments. Our code and web-based simulation tool are publicly available.<\/jats:p>","DOI":"10.3389\/fbinf.2026.1694775","type":"journal-article","created":{"date-parts":[[2026,3,16]],"date-time":"2026-03-16T06:40:46Z","timestamp":1773643246000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["Machine learning for N-dimensional spatial reasoning tasks on the web"],"prefix":"10.3389","volume":"6","author":[{"given":"Blake","family":"Moody","sequence":"first","affiliation":[{"name":"Department of Computer Science, University of Massachusetts Boston","place":["Boston, MA, United States"]}]},{"given":"JieHyun","family":"Kim","sequence":"additional","affiliation":[{"name":"Department of Computer Science, University of Massachusetts Boston","place":["Boston, MA, United States"]}]},{"given":"Sanghyuk","family":"Kim","sequence":"additional","affiliation":[{"name":"Department of Computer Science, University of Massachusetts Boston","place":["Boston, MA, United States"]}]},{"given":"Daniel","family":"Haehn","sequence":"additional","affiliation":[{"name":"Department of Computer Science, University of Massachusetts Boston","place":["Boston, MA, United States"]}]}],"member":"1965","published-online":{"date-parts":[[2026,3,16]]},"reference":[{"key":"B1","first-page":"2","article-title":"General n-dimensional rotations","volume-title":"WSCG SHORT communication papers proceedings","author":"Aguilera","year":"2004"},{"key":"B2","article-title":"Steerable equivariant representation learning","author":"Bhardwaj","year":"2023"},{"key":"B3","unstructured":"Threejs\n          \n          \n            \n              Cabello\n              R.\n            \n          \n          \n          2010"},{"key":"B4","volume-title":"A course in modern geometries","author":"Cederberg","year":"2010"},{"key":"B5","doi-asserted-by":"publisher","first-page":"15","DOI":"10.1007\/s41064-020-00102-3","article-title":"Geospatial artificial intelligence: potentials of machine learning for 3d point clouds and geospatial digital twins","volume":"88","author":"D\u00f6llner","year":"2020","journal-title":"PFG\u2013Journal Photogrammetry, Remote Sens. Geoinformation Sci."},{"key":"B6","volume-title":"ECMAScript language specification","year":"1997"},{"key":"B7","first-page":"3165","article-title":"Generalizing convolutional neural networks for equivariance to lie groups on arbitrary continuous data","volume-title":"International conference on machine learning (PMLR)","author":"Finzi","year":"2020"},{"key":"B8","doi-asserted-by":"publisher","first-page":"185","DOI":"10.1145\/3140587.3062363","article-title":"Bringing the web up to speed with webassembly","volume":"52","author":"Haas","year":"2017","journal-title":"SIGPLAN Not."},{"key":"B9","unstructured":"Web workers\n          \n          \n            \n              Hickson\n              I.\n            \n          \n          \n          2009"},{"key":"B10","doi-asserted-by":"publisher","first-page":"119","DOI":"10.1007\/s40708-016-0042-6","article-title":"Interactive machine learning for health informatics: when do we need the human-in-the-loop?","volume":"3","author":"Holzinger","year":"2016","journal-title":"Brain Informatics"},{"key":"B11","article-title":"Training machine learning models at the edge: a survey","author":"Khouas","year":"2024"},{"key":"B12","unstructured":"WebGL Specification, version 1.0\n          \n          \n          2011"},{"key":"B13","doi-asserted-by":"publisher","first-page":"43","DOI":"10.1109\/MSP.2017.2695801","article-title":"Optimal mass transport: signal processing and machine-learning applications","volume":"34","author":"Kolouri","year":"2017","journal-title":"IEEE Signal Processing Magazine"},{"key":"B14","doi-asserted-by":"publisher","first-page":"126","DOI":"10.3390\/info15030126","article-title":"Cyber security on the edge: efficient enabling of machine learning on iot devices","volume":"15","author":"Kumari","year":"2024","journal-title":"Information"},{"key":"B15","article-title":"Federated learning with communication delay in edge networks","author":"Lin","year":"2020","journal-title":"Corr. abs\/2008"},{"key":"B16","volume-title":"Robust machine learning models for high dimensional data interpretation","author":"Manigrasso","year":"2024"},{"key":"B17","doi-asserted-by":"publisher","first-page":"77","DOI":"10.1038\/s41597-023-01966-x","article-title":"Eco-isea3h, a machine learning ready spatial database for ecometric and species distribution modeling","volume":"10","author":"Mechenich","year":"2023","journal-title":"Sci. Data"},{"key":"B18","doi-asserted-by":"publisher","first-page":"20","DOI":"10.1007\/s44379-025-00016-0","article-title":"When physics meets machine learning: a survey of physics-informed machine learning","volume":"1","author":"Meng","year":"2025","journal-title":"Mach. Learn. Comput. Sci. Eng."},{"key":"B19","first-page":"23140","article-title":"Self-supervised learning of representations for space generates multi-modular grid cells","volume":"36","author":"Schaeffer","year":"2023","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"B20","first-page":"309","article-title":"Tensorflow. js: machine learning for the web and beyond","volume":"1","author":"Smilkov","year":"2019","journal-title":"Proc. Mach. Learn. Syst."},{"key":"B21","doi-asserted-by":"publisher","first-page":"35","DOI":"10.1016\/j.patrec.2024.11.010","article-title":"Ganzzle++: generative approaches for jigsaw puzzle solving as local to global assignment in latent spatial representations","volume":"187","author":"Talon","year":"2025","journal-title":"Pattern Recognit. Lett."},{"key":"B22","unstructured":"Webgpu specification\n          \n          \n          2021"},{"key":"B23","doi-asserted-by":"publisher","first-page":"2861","DOI":"10.1109\/TCYB.2019.2901897","article-title":"Reinforcement learning-based optimal sensor placement for spatiotemporal modeling","volume":"50","author":"Wang","year":"2019","journal-title":"IEEE Transactions Cybernetics"},{"key":"B24","doi-asserted-by":"publisher","first-page":"e38137","DOI":"10.1016\/j.heliyon.2024.e38137","article-title":"Federated learning: overview, strategies, applications, tools and future directions","volume":"10","author":"Yurdem","year":"2024","journal-title":"Heliyon"}],"container-title":["Frontiers in Bioinformatics"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fbinf.2026.1694775\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,3,16]],"date-time":"2026-03-16T06:40:49Z","timestamp":1773643249000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fbinf.2026.1694775\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,3,16]]},"references-count":24,"alternative-id":["10.3389\/fbinf.2026.1694775"],"URL":"https:\/\/doi.org\/10.3389\/fbinf.2026.1694775","relation":{},"ISSN":["2673-7647"],"issn-type":[{"value":"2673-7647","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,3,16]]},"article-number":"1694775"}}