[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"branding":3,"analytics":7,"article-rl-teaches-thinking-models-when-to-reason-not-how":10,"sections":41},{"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":36,"feedback":40,"feedback_at":22,"cost_usd":40,"total_tokens":40},4406,"rl-teaches-thinking-models-when-to-reason-not-how","RL Teaches Thinking Models When to Reason, Not How","A paper analyzing nine model pairs finds reinforcement learning teaches thinking models when to deploy reasoning skills the base model already has.","A new interpretability paper (arXiv:2510.07364) argues that the gap between a base language model and a \"thinking\" model is mostly about *when* to reason, not whether the model already knows *how*.\n\nThe researchers trained small Sparse Autoencoders on sentence-level activations drawn from model reasoning traces, generating what they call reasoning taxonomies. They then applied a technique called \"constructive model diffing\" to measure what training changed between a base model and its fine-tuned thinking counterpart. Across nine base\u002Fthinking model pairs (four trained with reinforcement learning, four with supervised fine-tuning distillation, and one mixed), two consistent findings emerged: RL-trained model taxonomies converged to far lower loss using base-model category vectors, and hybrid models recovered about 76% of the RL-to-thinking gap but only 11% of the SFT gap.\n\nThe split matters because it reframes what \"thinking\" models actually do differently. If RL training is largely teaching scheduling heuristics rather than new reasoning skills, then benchmarks that pit base models against thinking models may be measuring *when* the model chooses to reason, not how capable its reasoning is. The SFT-distillation finding cuts the other way: distilling a thinking model apparently installs genuinely new reasoning mechanisms, not just new heuristics for deploying old ones, which makes the SFT gap harder to close than the RL gap.\n\nThe paper is already on version four, which suggests it has faced meaningful scrutiny; treat the sparse autoencoder methodology as an active proposal in the interpretability community, not a settled technique.","[\"ai research\",\"language models\",\"reinforcement learning\",\"interpretability\"]","2026-07-08T04:00:00.000Z","2026-07-08T08:28:08.908Z","2026-07-08T08:28:11.716Z","published",null,[24],{"id":25,"reviewer":26,"round":27,"reason":28,"status":29},"editor-r1","editor",1,"The article omits author attribution and source identification (no author names, institution, or arXiv paper ID) required for the piece to be independently verifiable.","resolved","ai",[32,33,34,35],"ai research","language models","reinforcement learning","interpretability",[37],{"name":38,"url":39},"arXiv cs.AI","https:\u002F\u002Farxiv.org\u002Fabs\u002F2510.07364",0,{"sections":42},[43,47,52,57,62,67,72,77,82,87,92,96,101,106],{"name":44,"slug":30,"count":45,"latest_published_at":46},"AI",2590,"2026-07-16T04:00:00.000Z",{"name":48,"slug":49,"count":50,"latest_published_at":51},"Security","security",294,"2026-07-15T19:59:48.000Z",{"name":53,"slug":54,"count":55,"latest_published_at":56},"Deals","deals",179,"2026-06-29T20:02:07.000Z",{"name":58,"slug":59,"count":60,"latest_published_at":61},"Policy","policy",158,"2026-07-16T00:02:48.000Z",{"name":63,"slug":64,"count":65,"latest_published_at":66},"Hardware","hardware",122,"2026-07-14T19:46:26.000Z",{"name":68,"slug":69,"count":70,"latest_published_at":71},"Consumer Tech","consumer-tech",93,"2026-07-13T13:20:48.000Z",{"name":73,"slug":74,"count":75,"latest_published_at":76},"Software","software",70,"2026-07-13T19:52:25.000Z",{"name":78,"slug":79,"count":80,"latest_published_at":81},"Science","science",66,"2026-07-10T10:29:37.000Z",{"name":83,"slug":84,"count":85,"latest_published_at":86},"Dev Tools","dev-tools",59,"2026-07-07T04:00:00.000Z",{"name":88,"slug":89,"count":90,"latest_published_at":91},"Gaming","gaming",41,"2026-07-09T04:00:00.000Z",{"name":93,"slug":94,"count":90,"latest_published_at":95},"Startups","startups","2026-06-29T20:55:50.000Z",{"name":97,"slug":98,"count":99,"latest_published_at":100},"General","general",29,"2026-07-10T22:28:58.000Z",{"name":102,"slug":103,"count":104,"latest_published_at":105},"Reviews","reviews",20,"2026-06-24T12:00:01.000Z",{"name":107,"slug":108,"count":109,"latest_published_at":110},"How-To","how-to",6,"2026-06-16T09:00:00.000Z"]