[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"branding":3,"analytics":7,"article-demopsd-targets-a-core-flaw-in-ai-self-training":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},3553,"demopsd-targets-a-core-flaw-in-ai-self-training","DemoPSD Targets a Core Flaw in AI Self-Training","A new framework called DemoPSD fixes a shortcut-learning problem that quietly undermines how large language models train themselves to reason.","Self-training AI models have a cheating problem, and a new paper proposes a fix.\n\nWhen a large language model acts as its own teacher — a technique called on-policy self-distillation — it can accidentally pass hints to the student version of itself that won't exist at test time. The student learns to exploit those hints rather than reason from scratch, a flaw the researchers call privileged information leakage. DemoPSD, introduced in a preprint this week, addresses this by selectively filtering how much teacher guidance the student absorbs at each token position, using a blended target distribution that keeps the student's own reasoning intact. Tested across four scientific fields on the SciKnowEval benchmark, it outperformed both GRPO and SDPO while generalizing better to out-of-distribution questions on the GPQA benchmark.\n\nThe detail worth noting is what this reveals about the self-distillation trend itself. Training a single model to teach itself is appealing because it cuts costs and sidesteps the need for separate teacher models — but dense token-level supervision, the researchers argue, quietly trades generalization for in-domain pattern matching. DemoPSD's fix is not to abandon self-distillation but to make the student more skeptical of its teacher.\n\nThis lands in a crowded field of reasoning-improvement techniques — RLHF, chain-of-thought fine-tuning, process reward models — all competing to close the gap between benchmark performance and real-world reliability. Whether selectively ignoring your own teacher scales beyond scientific question-answering remains the obvious open question.","[\"ai\",\"machine-learning\",\"llm\",\"research\"]","2026-07-03T04:00:00.000Z","2026-07-03T08:30:56.128Z","2026-07-03T08:30:59.001Z","published",null,[],"ai",[24,26,27,28],"machine-learning","llm","research",[30],{"name":31,"url":32},"arXiv cs.AI","https:\u002F\u002Farxiv.org\u002Fabs\u002F2607.02502",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"]