[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"branding":3,"analytics":7,"article-a-post-hoc-patch-could-stop-ai-models-from-forgetting-old-skills":10,"sections":40},{"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":30,"tags":31,"sources":35,"feedback":39,"feedback_at":22,"cost_usd":39,"total_tokens":39},6877,"a-post-hoc-patch-could-stop-ai-models-from-forgetting-old-skills","A Post-Hoc Patch Could Stop AI Models From Forgetting Old Skills","A new post-hoc technique called JANUS restores forgotten skills in fine-tuned AI models without sacrificing accuracy on the new task it just learned.","AI models that get fine-tuned on new tasks tend to forget what they already knew, and a new paper proposes a fix that works after the fact rather than during training.\n\nThe paper introduces JANUS (JAcobian NUll Space), a post-hoc, tuning-agnostic weight-rectification framework. Existing parameter-efficient fine-tuning methods try to prevent this forgetting with a condition the authors call Subspace Orthogonality, which they argue is overly restrictive. JANUS instead projects parameter updates into a Jacobian null space to hit a different target, Parameter Space Orthogonality, which the authors say is the necessary and sufficient condition for preserving old performance to a first-order approximation. To keep that math valid beyond small updates, the method adds a Multi-step Adaptive Rectification step that checks a trust region and adjusts step size as it goes, plus efficiency tricks like ghost projection and sequence-level singular-value-decomposition compression.\n\nThat's a technical way of saying JANUS aims to solve the stability-plasticity dilemma properly: keep old-task accuracy intact without dragging down accuracy on the new task, rather than trading one for the other. Because it's post-hoc and tuning-agnostic, it can in principle sit on top of whatever fine-tuning method a team is already using, instead of requiring a new training recipe. The authors report it recovers historical knowledge across multiple fine-tuning methods while preserving downstream task performance.\n\nStill, these are the authors' own benchmarks in a preprint that hasn't been peer reviewed, so treat \"significantly mitigates\" as a claim to watch, not a verdict.","[\"ai\",\"fine-tuning\",\"catastrophic-forgetting\",\"machine-learning\"]","2026-09-18T04:00:00.000Z","2026-09-18T20:39:13.036Z","2026-09-18T20:39:24.929Z","published",null,[24],{"id":25,"reviewer":26,"round":27,"reason":28,"status":29},"editor-r1","editor",1,"The piece mischaracterizes the paper's stability-plasticity trade-off as being about training speed ('adapt fast, or remember well'), when the source is about whether new-task accuracy (plasticity) is preserved alongside old-task retention, not fine-tuning speed — rewrite the framing (including the dek's 'without slowing down new task learning') to reflect accuracy preservation, not speed.","resolved","ai",[30,32,33,34],"fine-tuning","catastrophic-forgetting","machine-learning",[36],{"name":37,"url":38},"arXiv cs.AI","https:\u002F\u002Farxiv.org\u002Fabs\u002F2609.19985",0,{"sections":41},[42,45,49,54,59,63,67,72,76,81,86,91,96,101],{"name":43,"slug":30,"count":44,"latest_published_at":18},"AI",4066,{"name":46,"slug":47,"count":48,"latest_published_at":18},"Security","security",657,{"name":50,"slug":51,"count":52,"latest_published_at":53},"Policy","policy",338,"2026-09-11T04:00:00.000Z",{"name":55,"slug":56,"count":57,"latest_published_at":58},"Deals","deals",179,"2026-06-29T20:02:07.000Z",{"name":60,"slug":61,"count":62,"latest_published_at":18},"Hardware","hardware",155,{"name":64,"slug":65,"count":66,"latest_published_at":18},"Science","science",123,{"name":68,"slug":69,"count":70,"latest_published_at":71},"Consumer Tech","consumer-tech",99,"2026-09-09T17:27:33.000Z",{"name":73,"slug":74,"count":75,"latest_published_at":18},"Dev Tools","dev-tools",78,{"name":77,"slug":78,"count":79,"latest_published_at":80},"Software","software",75,"2026-09-10T20:41:21.000Z",{"name":82,"slug":83,"count":84,"latest_published_at":85},"Startups","startups",55,"2026-09-09T23:14:29.000Z",{"name":87,"slug":88,"count":89,"latest_published_at":90},"Gaming","gaming",43,"2026-09-10T12:18:06.000Z",{"name":92,"slug":93,"count":94,"latest_published_at":95},"General","general",41,"2026-09-08T01:57:23.000Z",{"name":97,"slug":98,"count":99,"latest_published_at":100},"Reviews","reviews",20,"2026-06-24T12:00:01.000Z",{"name":102,"slug":103,"count":104,"latest_published_at":105},"How-To","how-to",6,"2026-06-16T09:00:00.000Z"]