[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"branding":3,"analytics":7,"article-a-fixed-version-of-muon-for-convolutions-works-no-better":10,"sections":34},{"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":29,"feedback":33,"feedback_at":22,"cost_usd":33,"total_tokens":33},10630,"a-fixed-version-of-muon-for-convolutions-works-no-better","A Fixed Version of Muon for Convolutions Works No Better","A team built a mathematically correct Muon optimizer for convolutions, and it performed no better than the shortcut version everyone already uses.","Muon, a popular optimizer for training neural networks, is built on math that doesn't actually apply to convolutional layers - and fixing that math didn't make it better.\n\nMuon works by computing a polar factor for matrix-shaped weight updates. Convolutional kernels are stored as four-dimensional tensors, so standard implementations just reshape them into matrices to use the same trick, even though that breaks the theoretical reasoning behind why Muon works. A group of researchers built an alternative called Conv-NS that does the polar-factor math directly in the actual convolutional geometry, preserving the kernel's structure instead of flattening it. They tested both versions on CIFAR-10 and ImageNet image classification and found the theoretically correct Conv-NS trained about as efficiently and landed on comparable accuracy to the reshape-based hack.\n\nThat's a strange result: the \"wrong\" method and the \"right\" method tie. The researchers' working theory is that forcing exact orthogonalization onto convolutional updates - the mathematically pure move - may actually overconstrain the optimizer rather than help it. It's a reminder that in deep learning, an optimizer's success is often empirical first, theory a distant second.\n\nCall it the Muon paradox: the shortcut survives the theory police and comes out even.","[\"ai\",\"machine-learning\",\"optimization\",\"research\"]","2026-10-07T04:00:00.000Z","2026-10-09T00:04:41.905Z","2026-10-09T00:04:42.167Z","published",null,[],"ai",[24,26,27,28],"machine-learning","optimization","research",[30],{"name":31,"url":32},"arXiv cs.AI","https:\u002F\u002Farxiv.org\u002Fabs\u002F2610.07103",0,{"sections":35},[36,40,44,49,54,58,62,67,72,77,81,86,91,96],{"name":37,"slug":24,"count":38,"latest_published_at":39},"AI",6778,"2026-10-09T04:00:00.000Z",{"name":41,"slug":42,"count":43,"latest_published_at":39},"Security","security",932,{"name":45,"slug":46,"count":47,"latest_published_at":48},"Policy","policy",486,"2026-10-08T22:40:11.000Z",{"name":50,"slug":51,"count":52,"latest_published_at":53},"Deals","deals",474,"2026-10-08T22:00:00.000Z",{"name":55,"slug":56,"count":57,"latest_published_at":39},"Hardware","hardware",232,{"name":59,"slug":60,"count":61,"latest_published_at":39},"Science","science",193,{"name":63,"slug":64,"count":65,"latest_published_at":66},"Consumer Tech","consumer-tech",181,"2026-10-08T23:26:35.000Z",{"name":68,"slug":69,"count":70,"latest_published_at":71},"Startups","startups",117,"2026-10-08T16:45:00.000Z",{"name":73,"slug":74,"count":75,"latest_published_at":76},"Software","software",114,"2026-10-08T17:57:01.000Z",{"name":78,"slug":79,"count":80,"latest_published_at":39},"Dev Tools","dev-tools",106,{"name":82,"slug":83,"count":84,"latest_published_at":85},"General","general",66,"2026-10-09T04:46:11.000Z",{"name":87,"slug":88,"count":89,"latest_published_at":90},"Gaming","gaming",58,"2026-10-08T20:08:45.000Z",{"name":92,"slug":93,"count":94,"latest_published_at":95},"Reviews","reviews",34,"2026-10-08T14:00:22.000Z",{"name":97,"slug":98,"count":99,"latest_published_at":100},"How-To","how-to",8,"2026-10-05T09:00:00.000Z"]