[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"branding":3,"analytics":7,"article-new-training-method-teaches-ai-models-to-reason-backward":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},7299,"new-training-method-teaches-ai-models-to-reason-backward","New Training Method Teaches AI Models to Reason Backward","A new method trains reasoning models to think backward as well as forward, aiming to fix five recurring math-reasoning failure modes.","A new paper argues today's top reasoning models struggle because they only know how to think in one direction.\n\nResearchers studied why models like GPT-o1, GPT-o3, and DeepSeek-R1 still botch math problems despite generating long chains of thought. Analyzing accuracy across multiple problem sets, they identified five recurring failure patterns: models don't cover enough of the solution space, they make computational mistakes, they run with unverified assumptions, they ignore stated constraints, and they simply hit the response-length limit before finishing. Their proposed fix is a backward reasoning pattern, trained using a two-stage easy-to-hard math dataset that pairs each problem with both a forward and a backward solution path. A fine-grained reward system with smoothed signals is meant to stop the model from gaming the reward, while a gradually shifting sampling schedule keeps forward and backward training examples balanced.\n\nChain-of-thought models have mostly been improved by making them think longer, not differently - more tokens, more steps, same direction every time. This work treats reasoning direction as a choice the model should make for itself, which is a cheaper lever to pull than just throwing more test-time compute at a problem.\n\nThe paper reports improved accuracy and efficiency on math proofs, but it doesn't publish head-to-head numbers against the models it critiques, so the real test is whether this holds up outside the benchmark built to showcase it.","[\"reasoning-models\",\"chain-of-thought\",\"llm-training\",\"math-ai\"]","2026-09-23T04:00:00.000Z","2026-09-23T06:01:19.554Z","2026-09-23T06:01:26.019Z","published",null,[],"ai",[26,27,28,29],"reasoning-models","chain-of-thought","llm-training","math-ai",[31],{"name":32,"url":33},"arXiv cs.AI","https:\u002F\u002Farxiv.org\u002Fabs\u002F2609.24760",0,{"sections":36},[37,40,44,49,54,59,63,68,73,78,83,88,93,98],{"name":38,"slug":24,"count":39,"latest_published_at":18},"AI",4264,{"name":41,"slug":42,"count":43,"latest_published_at":18},"Security","security",707,{"name":45,"slug":46,"count":47,"latest_published_at":48},"Policy","policy",369,"2026-09-23T02:13:52.000Z",{"name":50,"slug":51,"count":52,"latest_published_at":53},"Deals","deals",202,"2026-09-22T23:00:04.000Z",{"name":55,"slug":56,"count":57,"latest_published_at":58},"Hardware","hardware",168,"2026-09-22T23:56:03.000Z",{"name":60,"slug":61,"count":62,"latest_published_at":18},"Science","science",133,{"name":64,"slug":65,"count":66,"latest_published_at":67},"Consumer Tech","consumer-tech",110,"2026-09-22T20:00:00.000Z",{"name":69,"slug":70,"count":71,"latest_published_at":72},"Software","software",80,"2026-09-22T23:32:52.000Z",{"name":74,"slug":75,"count":76,"latest_published_at":77},"Dev Tools","dev-tools",79,"2026-09-22T22:21:13.000Z",{"name":79,"slug":80,"count":81,"latest_published_at":82},"Startups","startups",65,"2026-09-22T22:06:48.000Z",{"name":84,"slug":85,"count":86,"latest_published_at":87},"Gaming","gaming",45,"2026-09-22T15:35:06.000Z",{"name":89,"slug":90,"count":91,"latest_published_at":92},"General","general",43,"2026-09-21T23:48:56.000Z",{"name":94,"slug":95,"count":96,"latest_published_at":97},"Reviews","reviews",27,"2026-09-22T13:00:00.000Z",{"name":99,"slug":100,"count":101,"latest_published_at":102},"How-To","how-to",6,"2026-06-16T09:00:00.000Z"]