[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"branding":3,"analytics":7,"article-researchers-cast-doubt-on-a-popular-ai-alignment-technique":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},6908,"researchers-cast-doubt-on-a-popular-ai-alignment-technique","Researchers Cast Doubt on a Popular AI Alignment Technique","A new study tests whether feeding models alignment-relevant text during pretraining actually makes them safer, and finds the effect breaks easily.","A new paper finds little proof that a popular AI safety technique actually works.\n\nThe technique is called alignment midtraining, or AMT: instead of only fine-tuning a model at the end of training, labs feed it large volumes of alignment-relevant text earlier, during pretraining, hoping the behavior generalizes better once the model is deployed. Researchers tested the idea on models up to 110 billion parameters, using as much as 1 billion tokens of midtraining data. They found midtraining could nudge a model's underlying \"motivation\" in simple test cases. But that effect vanished the moment even a tiny fraction of later fine-tuning data hinted at a different motivation. In a separate test involving rule-following, models only reliably learned a rule if it showed up somewhere in the midtraining or post-training data - implicit generalization to unstated rules mostly didn't happen.\n\nThis matters because midtraining has been floated as a way to make alignment hold up in situations a model was never explicitly trained for - which is most of deployment. If a handful of contradictory examples in fine-tuning can undo it, and rules still need to be spelled out rather than inferred, midtraining looks less like a generalization breakthrough and more like another layer of pattern-matching with the same blind spots as standard fine-tuning.\n\nThe authors are blunt about the implication: they say there isn't enough public evidence to claim midtraining solves the core problem of aligning increasingly capable systems. For a field that sometimes talks about alignment techniques as though they're closer to solved than they are, a result this easy to break is worth sitting with.","[\"ai-safety\",\"alignment\",\"llm-research\",\"pretraining\"]","2026-09-18T04:00:00.000Z","2026-09-18T22:05:12.809Z","2026-09-18T22:05:24.801Z","published",null,[],"ai",[26,27,28,29],"ai-safety","alignment","llm-research","pretraining",[31],{"name":32,"url":33},"arXiv cs.AI","https:\u002F\u002Farxiv.org\u002Fabs\u002F2609.20412",0,{"sections":36},[37,40,44,49,54,58,62,67,71,76,81,86,91,96],{"name":38,"slug":24,"count":39,"latest_published_at":18},"AI",4082,{"name":41,"slug":42,"count":43,"latest_published_at":18},"Security","security",661,{"name":45,"slug":46,"count":47,"latest_published_at":48},"Policy","policy",339,"2026-09-17T12:00:00.000Z",{"name":50,"slug":51,"count":52,"latest_published_at":53},"Deals","deals",179,"2026-06-29T20:02:07.000Z",{"name":55,"slug":56,"count":57,"latest_published_at":18},"Hardware","hardware",155,{"name":59,"slug":60,"count":61,"latest_published_at":18},"Science","science",125,{"name":63,"slug":64,"count":65,"latest_published_at":66},"Consumer Tech","consumer-tech",99,"2026-09-09T17:27:33.000Z",{"name":68,"slug":69,"count":70,"latest_published_at":18},"Dev Tools","dev-tools",78,{"name":72,"slug":73,"count":74,"latest_published_at":75},"Software","software",75,"2026-09-10T20:41:21.000Z",{"name":77,"slug":78,"count":79,"latest_published_at":80},"Startups","startups",55,"2026-09-09T23:14:29.000Z",{"name":82,"slug":83,"count":84,"latest_published_at":85},"Gaming","gaming",43,"2026-09-10T12:18:06.000Z",{"name":87,"slug":88,"count":89,"latest_published_at":90},"General","general",41,"2026-09-08T01:57:23.000Z",{"name":92,"slug":93,"count":94,"latest_published_at":95},"Reviews","reviews",20,"2026-06-24T12:00:01.000Z",{"name":97,"slug":98,"count":99,"latest_published_at":100},"How-To","how-to",6,"2026-06-16T09:00:00.000Z"]