[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"branding":3,"analytics":7,"article-a-training-free-way-to-combine-lora-experts-in-ai-models":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},8193,"a-training-free-way-to-combine-lora-experts-in-ai-models","A Training-Free Way to Combine LoRA Experts in AI Models","A new framework called JevSoup picks two LoRA experts per query and blends them mathematically, skipping the training routers usually require.","A new routing trick lets AI systems pick and combine specialized skills without any extra training.\n\nLoRA adapters are small, swappable modules that give a base model a narrow skill, and developers increasingly stack many of them together like a toolbox. The problem has been picking the right tools for a given task without training a separate router or feeding it extra data. JevSoup splits that job into a fast \"System One\" step that selects experts and a slower \"System Two\" step that runs them, choosing two adapters using nothing but the input text and each expert's written description. It keeps the top-ranked expert's update in full, mathematically strips out the overlap from the second expert so the two do not cancel each other out, then merges them with equal weight. Across 14 PorTAL benchmark tasks and three sizes of the Qwen3 model, this got accuracy gains of up to 1.19 percent (task-macro) and 1.21 percent (sample-micro) over the best existing baselines, with code posted on GitHub.\n\nThe real story here is cost, not accuracy. Most adapter-routing systems require training a router network or collecting auxiliary data every time someone adds a new skill module, which does not scale as adapter libraries grow into the hundreds. Treating routing as a read-the-descriptions problem instead means new experts can be dropped in with zero retraining, which matters far more once teams are managing dozens of LoRA modules than it does in a two-adapter demo.\n\nA sub-1.5 percent accuracy bump is not nothing, but it is also not the headline - the headline is that you get it for free.","[\"lora\",\"model-routing\",\"fine-tuning\",\"mixture-of-experts\"]","2026-09-28T04:00:00.000Z","2026-09-28T16:19:17.242Z","2026-09-28T16:19:24.027Z","published",null,[],"ai",[26,27,28,29],"lora","model-routing","fine-tuning","mixture-of-experts",[31],{"name":32,"url":33},"arXiv cs.AI","https:\u002F\u002Farxiv.org\u002Fabs\u002F2609.30922",0,{"sections":36},[37,40,44,49,54,59,63,68,73,78,83,88,92,97],{"name":38,"slug":24,"count":39,"latest_published_at":18},"AI",4844,{"name":41,"slug":42,"count":43,"latest_published_at":18},"Security","security",762,{"name":45,"slug":46,"count":47,"latest_published_at":48},"Policy","policy",399,"2026-09-27T18:39:02.000Z",{"name":50,"slug":51,"count":52,"latest_published_at":53},"Deals","deals",266,"2026-09-28T14:00:00.000Z",{"name":55,"slug":56,"count":57,"latest_published_at":58},"Hardware","hardware",189,"2026-09-28T10:52:40.000Z",{"name":60,"slug":61,"count":62,"latest_published_at":18},"Science","science",151,{"name":64,"slug":65,"count":66,"latest_published_at":67},"Consumer Tech","consumer-tech",135,"2026-09-26T14:30:00.000Z",{"name":69,"slug":70,"count":71,"latest_published_at":72},"Software","software",91,"2026-09-25T20:55:00.000Z",{"name":74,"slug":75,"count":76,"latest_published_at":77},"Dev Tools","dev-tools",84,"2026-09-26T04:20:58.000Z",{"name":79,"slug":80,"count":81,"latest_published_at":82},"Startups","startups",76,"2026-09-25T18:33:59.000Z",{"name":84,"slug":85,"count":86,"latest_published_at":87},"Gaming","gaming",48,"2026-09-25T18:35:21.000Z",{"name":89,"slug":90,"count":86,"latest_published_at":91},"General","general","2026-09-26T17:02:42.000Z",{"name":93,"slug":94,"count":95,"latest_published_at":96},"Reviews","reviews",30,"2026-09-24T20:07:31.000Z",{"name":98,"slug":99,"count":100,"latest_published_at":101},"How-To","how-to",6,"2026-06-16T09:00:00.000Z"]