[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"branding":3,"analytics":7,"article-ai-workflow-generator-skips-retraining-for-every-tradeoff":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},8856,"ai-workflow-generator-skips-retraining-for-every-tradeoff","AI Workflow Generator Skips Retraining for Every Tradeoff","MoFlow searches once and returns an AI workflow tuned to any mix of accuracy, cost, and latency, no retraining required.","MoFlow is a new method for building AI agent workflows that can be retuned for accuracy, cost, or speed without retraining from scratch.\n\nResearchers behind the work frame workflow generation as a multi-objective decision process, searched with a technique called Convex-Hull Monte Carlo Tree Search with optimistic set-valued backups. Instead of each node tracking a single weighted score, it stores a set of reachable tradeoffs. One search run approximates the full range of good options, known as the Pareto front, so MoFlow can hand back a workflow matching any stated preference by lookup rather than rerunning the whole search. The team tested it against six baseline methods across six benchmarks covering math, code, and question answering. Because those baselines only optimize a single number, the comparison was set up in their favor: baselines were rerun fresh for every test preference, while MoFlow never saw those preferences in advance. MoFlow still posted the highest average hypervolume, a standard measure of how well a set of tradeoffs covers the space of good outcomes.\n\nMost agentic workflow generators lock in one tradeoff at training time. Want a cheaper version for a cost-sensitive deployment, or a faster one for a latency-sensitive product? Normally that means retraining or re-searching from zero. MoFlow's pitch is that one search session covers the whole menu of tradeoffs, which matters for any team iterating on workflow cost and performance after launch, not just at model-selection time.\n\nThat said, this is a preprint, not a published, peer-reviewed result, and winning on hypervolume across six benchmarks is a research-paper victory, not evidence it survives contact with messier production pipelines.","[\"ai\",\"agentic-workflows\",\"ai-research\",\"dev-tools\"]","2026-10-01T04:00:00.000Z","2026-10-01T07:44:10.493Z","2026-10-01T07:44:15.673Z","published",null,[],"ai",[24,26,27,28],"agentic-workflows","ai-research","dev-tools",[30],{"name":31,"url":32},"arXiv cs.AI","https:\u002F\u002Farxiv.org\u002Fabs\u002F2609.38294",0,{"sections":35},[36,39,44,49,54,59,64,69,73,77,82,87,92,97],{"name":37,"slug":24,"count":38,"latest_published_at":18},"AI",5270,{"name":40,"slug":41,"count":42,"latest_published_at":43},"Security","security",801,"2026-09-30T22:18:23.000Z",{"name":45,"slug":46,"count":47,"latest_published_at":48},"Policy","policy",429,"2026-10-01T02:26:17.000Z",{"name":50,"slug":51,"count":52,"latest_published_at":53},"Deals","deals",298,"2026-09-30T21:00:26.000Z",{"name":55,"slug":56,"count":57,"latest_published_at":58},"Hardware","hardware",196,"2026-09-30T13:00:00.000Z",{"name":60,"slug":61,"count":62,"latest_published_at":63},"Science","science",157,"2026-09-30T15:00:56.000Z",{"name":65,"slug":66,"count":67,"latest_published_at":68},"Consumer Tech","consumer-tech",149,"2026-09-30T22:57:11.000Z",{"name":70,"slug":28,"count":71,"latest_published_at":72},"Dev Tools",93,"2026-10-01T02:30:48.000Z",{"name":74,"slug":75,"count":71,"latest_published_at":76},"Software","software","2026-09-30T21:41:11.000Z",{"name":78,"slug":79,"count":80,"latest_published_at":81},"Startups","startups",84,"2026-09-30T20:39:09.000Z",{"name":83,"slug":84,"count":85,"latest_published_at":86},"Gaming","gaming",51,"2026-09-30T16:24:30.000Z",{"name":88,"slug":89,"count":90,"latest_published_at":91},"General","general",50,"2026-09-30T21:37:54.000Z",{"name":93,"slug":94,"count":95,"latest_published_at":96},"Reviews","reviews",31,"2026-09-28T14:31:34.000Z",{"name":98,"slug":99,"count":100,"latest_published_at":101},"How-To","how-to",6,"2026-06-16T09:00:00.000Z"]