[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"branding":3,"analytics":7,"article-a-training-trick-teaches-ai-models-when-to-stop-overthinking":10,"sections":34},{"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":29,"feedback":33,"feedback_at":22,"cost_usd":33,"total_tokens":33},6774,"a-training-trick-teaches-ai-models-when-to-stop-overthinking","A Training Trick Teaches AI Models When to Stop Overthinking","When2Think teaches reasoning models to skip lengthy reasoning on easy problems, cutting tokens while raising accuracy on math benchmarks.","Researchers have built a training method that teaches AI reasoning models when to think hard and when not to bother.\n\nThe technique, called When2Think, is a post-training framework described in a new arXiv paper. It targets \"hybrid\" reasoning models, systems that can either answer directly or work through a problem step by step before answering. Today's models tend to get this allocation wrong: they burn tokens over-explaining simple questions and cut corners on genuinely hard ones. When2Think uses a reward-shaping method called Instance-level Difficulty-Aware Control, which draws on pre-computed accuracy and token-usage statistics to nudge the model toward quick answers on easy problems and extended reasoning on hard ones, without needing a separate learned reward model.\n\nReasoning models are expensive to run precisely because they generate long chains of intermediate steps, and most efficiency fixes so far have applied a blunt, uniform penalty on length or a rigid router, trading accuracy on hard problems for savings on easy ones. When2Think's difficulty-aware approach reports the opposite trade: on the AIME24 math benchmark, Pass@3 accuracy rose 10.0 percent while token usage fell 27.9 percent versus the base model, and it hit 40.0 percent Pass@3 on the newer AIME25 set, beating both compression and routing-only baselines.\n\nThe gains are on math contest problems, a narrow and heavily benchmarked domain, so the real test is whether difficulty-aware routing holds up on messier, real-world tasks.","[\"ai\",\"reasoning-models\",\"llm-efficiency\",\"arxiv\"]","2026-09-18T04:00:00.000Z","2026-09-18T16:05:23.090Z","2026-09-18T16:05:34.995Z","published",null,[],"ai",[24,26,27,28],"reasoning-models","llm-efficiency","arxiv",[30],{"name":31,"url":32},"arXiv cs.AI","https:\u002F\u002Farxiv.org\u002Fabs\u002F2609.19671",0,{"sections":35},[36,39,43,48,53,57,61,66,71,76,81,86,91,96],{"name":37,"slug":24,"count":38,"latest_published_at":18},"AI",3958,{"name":40,"slug":41,"count":42,"latest_published_at":18},"Security","security",652,{"name":44,"slug":45,"count":46,"latest_published_at":47},"Policy","policy",338,"2026-09-11T04:00:00.000Z",{"name":49,"slug":50,"count":51,"latest_published_at":52},"Deals","deals",179,"2026-06-29T20:02:07.000Z",{"name":54,"slug":55,"count":56,"latest_published_at":18},"Hardware","hardware",155,{"name":58,"slug":59,"count":60,"latest_published_at":18},"Science","science",116,{"name":62,"slug":63,"count":64,"latest_published_at":65},"Consumer Tech","consumer-tech",99,"2026-09-09T17:27:33.000Z",{"name":67,"slug":68,"count":69,"latest_published_at":70},"Dev Tools","dev-tools",76,"2026-09-18T01:04:54.000Z",{"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"]