[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"branding":3,"analytics":7,"article-tinycast-squeezes-forecasting-ai-into-146k-parameters":10,"sections":35},{"siteName":4,"siteTagline":5,"publisherName":4,"contactEmail":6},"The Revision","Tech news, decoded.","editor@therevision.news",{"gaMeasurementId":8,"adsenseClientId":9},"G-ZW2MV82GYR","ca-pub-8533917693782264",{"article":11},{"id":12,"slug":13,"title":14,"dek":15,"body_md":16,"tags_json":17,"published_at":18,"created_at":19,"updated_at":20,"status":21,"review_note":22,"review_notes":23,"image_url":22,"persona_id":22,"persona_name":22,"section":24,"tags":25,"sources":30,"feedback":34,"feedback_at":22,"cost_usd":34,"total_tokens":34},5399,"tinycast-squeezes-forecasting-ai-into-146k-parameters","TinyCast Squeezes Forecasting AI Into 146K Parameters","A new zero-shot forecasting model skips learning periodicity and computes it instead, beating far larger rivals while running on embedded hardware.","A new forecasting model gets its edge by refusing to learn something it can just calculate.\n\nTinyCast is a 146,505-parameter zero-shot forecaster described in a new arXiv paper. Instead of learning periodic patterns from training data, it uses a zero-parameter spectral detector to find the dominant cycles in a given data series, then folds the input on that phase before a small convolutional encoder and quantile decoder handle the rest. On the GIFT-Eval benchmark, it is smaller than every other zero-shot model whose parameter count is known, and it reportedly sets the size-to-accuracy frontier for probabilistic forecasts. On the Chronos-ZS and fev-bench benchmarks, every neural model that outperforms it is at least 28 times its size.\n\nThe interesting move here isn't the accuracy claim, it's the architecture bet: treat periodicity as a computable property of the input, not a pattern to be memorized. That sidesteps the usual scaling logic in forecasting research, where bigger context windows and more parameters are assumed to buy better predictions. It also matters for where forecasting can run. Because the model's math is limited to convolutions and matrix multiplications, it exports to static INT8 and can run entirely on an embedded device without fitting itself to each new signal first.\n\nMost \"zero-shot\" forecasting claims still assume a GPU is around. A model this small forecasting demand curves or sensor drift on a microcontroller is the more practical story, if the benchmark numbers hold up outside the paper's own test suite.","[\"forecasting\",\"zero-shot-learning\",\"embedded-ai\",\"machine-learning\"]","2026-08-18T04:00:00.000Z","2026-08-18T18:38:12.402Z","2026-08-18T18:38:24.396Z","published",null,[],"ai",[26,27,28,29],"forecasting","zero-shot-learning","embedded-ai","machine-learning",[31],{"name":32,"url":33},"arXiv cs.AI","https:\u002F\u002Farxiv.org\u002Fabs\u002F2608.15767",0,{"sections":36},[37,41,45,50,55,60,65,70,75,79,84,89,94,99],{"name":38,"slug":24,"count":39,"latest_published_at":40},"AI",3293,"2026-08-20T04:00:00.000Z",{"name":42,"slug":43,"count":44,"latest_published_at":40},"Security","security",435,{"name":46,"slug":47,"count":48,"latest_published_at":49},"Policy","policy",210,"2026-08-19T09:32:27.000Z",{"name":51,"slug":52,"count":53,"latest_published_at":54},"Deals","deals",179,"2026-06-29T20:02:07.000Z",{"name":56,"slug":57,"count":58,"latest_published_at":59},"Hardware","hardware",140,"2026-08-19T18:25:42.000Z",{"name":61,"slug":62,"count":63,"latest_published_at":64},"Consumer Tech","consumer-tech",95,"2026-08-18T16:05:00.000Z",{"name":66,"slug":67,"count":68,"latest_published_at":69},"Science","science",90,"2026-08-19T18:41:02.000Z",{"name":71,"slug":72,"count":73,"latest_published_at":74},"Software","software",73,"2026-08-18T07:51:50.000Z",{"name":76,"slug":77,"count":78,"latest_published_at":18},"Dev Tools","dev-tools",69,{"name":80,"slug":81,"count":82,"latest_published_at":83},"Startups","startups",47,"2026-08-19T19:13:46.000Z",{"name":85,"slug":86,"count":87,"latest_published_at":88},"Gaming","gaming",41,"2026-07-09T04:00:00.000Z",{"name":90,"slug":91,"count":92,"latest_published_at":93},"General","general",33,"2026-08-18T22:18:13.000Z",{"name":95,"slug":96,"count":97,"latest_published_at":98},"Reviews","reviews",20,"2026-06-24T12:00:01.000Z",{"name":100,"slug":101,"count":102,"latest_published_at":103},"How-To","how-to",6,"2026-06-16T09:00:00.000Z"]