[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"branding":3,"analytics":7,"article-spark-steers-llms-from-inside-to-fix-reasoning-failures":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},4667,"spark-steers-llms-from-inside-to-fix-reasoning-failures","SPARK Steers LLMs From Inside to Fix Reasoning Failures","A new method reads hidden model states to diagnose why an LLM got an answer wrong, then nudges it toward better reasoning without retraining.","Researchers have a new way to fix LLM reasoning that skips the usual guesswork and goes straight to the model's internals.\n\nThe technique, called SPARK, monitors hidden states inside a frozen large language model to detect when the model fails to activate reasoning it's already capable of. Rather than judging a model purely on whether its final answer was right or wrong, SPARK diagnoses the cause — distinguishing between a missing capability and a reasoning process that misfired. It then applies lightweight steering at test time, requiring no retraining or fine-tuning. A key finding: raw hidden-state signals are skewed by prompt length, which matters especially for math and code problems where harder questions are simply longer. SPARK controls for that before deciding where to intervene.\n\nMost LLM evaluation methods stop at the output layer, which tells you what went wrong but not why. SPARK's internal profiling makes the failure mode legible — which opens a path to targeted fixes rather than blanket prompting tricks. On MATH-500, the method lifted accuracy for Qwen3-4B from 82.0% to 84.6% and for Qwen3-8B from 82.4% to 85.6%, modest but consistent gains with no weight updates.\n\nThe gains are real but narrow — a few percentage points on models already clearing 82% — and the method has only been shown on Qwen3 variants so far, leaving broader generalization an open question.","[\"ai\",\"llm\",\"reasoning\",\"research\"]","2026-07-14T04:00:00.000Z","2026-07-14T05:03:27.367Z","2026-07-14T05:03:30.208Z","published",null,[],"ai",[24,26,27,28],"llm","reasoning","research",[30],{"name":31,"url":32},"arXiv cs.AI","https:\u002F\u002Farxiv.org\u002Fabs\u002F2607.10296",0,{"sections":35},[36,40,45,50,55,60,65,70,75,80,85,89,94,99],{"name":37,"slug":24,"count":38,"latest_published_at":39},"AI",2590,"2026-07-16T04:00:00.000Z",{"name":41,"slug":42,"count":43,"latest_published_at":44},"Security","security",294,"2026-07-15T19:59:48.000Z",{"name":46,"slug":47,"count":48,"latest_published_at":49},"Deals","deals",179,"2026-06-29T20:02:07.000Z",{"name":51,"slug":52,"count":53,"latest_published_at":54},"Policy","policy",158,"2026-07-16T00:02:48.000Z",{"name":56,"slug":57,"count":58,"latest_published_at":59},"Hardware","hardware",122,"2026-07-14T19:46:26.000Z",{"name":61,"slug":62,"count":63,"latest_published_at":64},"Consumer Tech","consumer-tech",93,"2026-07-13T13:20:48.000Z",{"name":66,"slug":67,"count":68,"latest_published_at":69},"Software","software",70,"2026-07-13T19:52:25.000Z",{"name":71,"slug":72,"count":73,"latest_published_at":74},"Science","science",66,"2026-07-10T10:29:37.000Z",{"name":76,"slug":77,"count":78,"latest_published_at":79},"Dev Tools","dev-tools",59,"2026-07-07T04:00:00.000Z",{"name":81,"slug":82,"count":83,"latest_published_at":84},"Gaming","gaming",41,"2026-07-09T04:00:00.000Z",{"name":86,"slug":87,"count":83,"latest_published_at":88},"Startups","startups","2026-06-29T20:55:50.000Z",{"name":90,"slug":91,"count":92,"latest_published_at":93},"General","general",29,"2026-07-10T22:28:58.000Z",{"name":95,"slug":96,"count":97,"latest_published_at":98},"Reviews","reviews",20,"2026-06-24T12:00:01.000Z",{"name":100,"slug":101,"count":102,"latest_published_at":103},"How-To","how-to",6,"2026-06-16T09:00:00.000Z"]