[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"branding":3,"analytics":7,"article-a-fix-for-noisy-codebooks-in-hyperdimensional-computing":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},9020,"a-fix-for-noisy-codebooks-in-hyperdimensional-computing","A Fix for Noisy Codebooks in Hyperdimensional Computing","A new paper shows how to build hyperdimensional computing codebooks that hit their target similarity curve exactly, instead of hoping randomness gets close.","Hyperdimensional computing just got a more precise way to turn numbers into vectors.\n\nResearchers published a framework for building \"scalar codebooks\" - the lookup tables that represent quantized numeric values as dense binary hypervectors - so the similarity between any two codes tracks a target mathematical curve exactly, rather than approximately. Today's codebooks are typically built by randomly flipping bits, a method that only behaves as intended once dimensionality gets very large. At the lower dimensions real hardware actually uses, sampling noise, lopsided starting points, and uneven bit updates make the codebook drift from the curve it is supposed to follow. The paper works out deterministic constraints on starting bit counts, update frequency, and update balance that force the codebook to match either an exponential or linear similarity decay, and backs this with exact formulas for bias, variance, and error, confirmed through simulations across multiple vector sizes and quantization ranges.\n\nHyperdimensional computing is pitched as a lightweight alternative to deep learning for sensor and edge devices, but its accuracy depends on these codebooks behaving the way the underlying math assumes. Closing the gap between theory and practice matters most exactly where HDC is supposed to shine: low-power chips that cannot afford thousands of dimensions to average away randomness.\n\nThis is not a product or a benchmark win. It is plumbing work that makes a decade-old technique a bit more trustworthy at the dimensions people actually ship.","[\"hyperdimensional-computing\",\"vector-symbolic-architectures\",\"ai-research\",\"edge-ai\"]","2026-10-01T04:00:00.000Z","2026-10-01T16:03:56.744Z","2026-10-01T16:04:00.985Z","published",null,[],"ai",[26,27,28,29],"hyperdimensional-computing","vector-symbolic-architectures","ai-research","edge-ai",[31],{"name":32,"url":33},"arXiv cs.AI","https:\u002F\u002Farxiv.org\u002Fabs\u002F2609.38471",0,{"sections":36},[37,40,44,49,54,59,63,68,73,77,82,87,92,97],{"name":38,"slug":24,"count":39,"latest_published_at":18},"AI",5487,{"name":41,"slug":42,"count":43,"latest_published_at":18},"Security","security",809,{"name":45,"slug":46,"count":47,"latest_published_at":48},"Policy","policy",429,"2026-10-01T02:26:17.000Z",{"name":50,"slug":51,"count":52,"latest_published_at":53},"Deals","deals",298,"2026-09-30T21:00:26.000Z",{"name":55,"slug":56,"count":57,"latest_published_at":58},"Hardware","hardware",196,"2026-09-30T13:00:00.000Z",{"name":60,"slug":61,"count":62,"latest_published_at":18},"Science","science",162,{"name":64,"slug":65,"count":66,"latest_published_at":67},"Consumer Tech","consumer-tech",149,"2026-09-30T22:57:11.000Z",{"name":69,"slug":70,"count":71,"latest_published_at":72},"Dev Tools","dev-tools",93,"2026-10-01T02:30:48.000Z",{"name":74,"slug":75,"count":71,"latest_published_at":76},"Software","software","2026-09-30T21:41:11.000Z",{"name":78,"slug":79,"count":80,"latest_published_at":81},"Startups","startups",84,"2026-09-30T20:39:09.000Z",{"name":83,"slug":84,"count":85,"latest_published_at":86},"Gaming","gaming",51,"2026-09-30T16:24:30.000Z",{"name":88,"slug":89,"count":90,"latest_published_at":91},"General","general",50,"2026-09-30T21:37:54.000Z",{"name":93,"slug":94,"count":95,"latest_published_at":96},"Reviews","reviews",31,"2026-09-28T14:31:34.000Z",{"name":98,"slug":99,"count":100,"latest_published_at":101},"How-To","how-to",6,"2026-06-16T09:00:00.000Z"]