[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"branding":3,"analytics":7,"article-a-cheaper-way-to-change-what-ai-models-believe":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},9667,"a-cheaper-way-to-change-what-ai-models-believe","A Cheaper Way to Change What AI Models Believe","A new technique called grafting edits AI beliefs without full retraining, and fixes a glitch where edits bled into unrelated facts.","Researchers have found a cheaper way to change what an AI model believes, without retraining it from scratch.\n\nThe standard method, called synthetic document fine-tuning, feeds a model fabricated documents to implant a belief, such as a fake fact or a personality trait. Done early, during pre-training, it works cleanly, but pre-training runs are slow and expensive to redo, so most labs apply the technique later, to an already post-trained model. That shortcut has a cost: the model starts treating unrelated fabricated entities as real too, a glitch the researchers call reality drift, and it degrades the model's existing capabilities. Their fix, called grafting, trains the belief-change adapter on the original pre-trained checkpoint, then adds that weight update onto the finished, post-trained model.\n\nTested on model families up to 284 billion parameters, including attempts to implant false facts and build intentionally misaligned test models, grafting matched the standard method's ability to install a belief while cutting reality drift and loss of coherence by more than half. That matters because alignment research runs on iteration: right now, every tweak to a model's early training requires redoing the entire expensive post-training process just to check if it worked. A technique that skips that step turns months of testing into a single fine-tuning run.\n\nIt is a workaround, not a cure: grafting approximates what a clean pre-training run would produce, it does not replace one.","[\"ai-alignment\",\"model-training\",\"ai-safety\",\"research\"]","2026-10-02T04:00:00.000Z","2026-10-03T05:20:03.279Z","2026-10-03T05:20:08.521Z","published",null,[],"ai",[26,27,28,29],"ai-alignment","model-training","ai-safety","research",[31],{"name":32,"url":33},"arXiv cs.AI","https:\u002F\u002Farxiv.org\u002Fabs\u002F2610.00767",0,{"sections":36},[37,40,44,48,53,57,61,66,71,76,81,86,91,96],{"name":38,"slug":24,"count":39,"latest_published_at":18},"AI",5977,{"name":41,"slug":42,"count":43,"latest_published_at":18},"Security","security",842,{"name":45,"slug":46,"count":47,"latest_published_at":18},"Policy","policy",438,{"name":49,"slug":50,"count":51,"latest_published_at":52},"Deals","deals",317,"2026-10-01T22:00:00.000Z",{"name":54,"slug":55,"count":56,"latest_published_at":18},"Hardware","hardware",199,{"name":58,"slug":59,"count":60,"latest_published_at":18},"Science","science",173,{"name":62,"slug":63,"count":64,"latest_published_at":65},"Consumer Tech","consumer-tech",155,"2026-10-01T19:54:10.000Z",{"name":67,"slug":68,"count":69,"latest_published_at":70},"Dev Tools","dev-tools",96,"2026-10-01T16:57:03.000Z",{"name":72,"slug":73,"count":74,"latest_published_at":75},"Software","software",93,"2026-09-30T21:41:11.000Z",{"name":77,"slug":78,"count":79,"latest_published_at":80},"Startups","startups",90,"2026-10-01T21:55:22.000Z",{"name":82,"slug":83,"count":84,"latest_published_at":85},"Gaming","gaming",53,"2026-10-02T02:50:39.000Z",{"name":87,"slug":88,"count":89,"latest_published_at":90},"General","general",50,"2026-09-30T21:37:54.000Z",{"name":92,"slug":93,"count":94,"latest_published_at":95},"Reviews","reviews",31,"2026-09-28T14:31:34.000Z",{"name":97,"slug":98,"count":99,"latest_published_at":100},"How-To","how-to",7,"2026-10-01T09:00:00.000Z"]