[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"branding":3,"analytics":7,"article-a-nav-fix-for-vln-agents-that-gets-lost-or-goes-in-circles":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},3285,"a-nav-fix-for-vln-agents-that-gets-lost-or-goes-in-circles","A Nav Fix for VLN Agents That Gets Lost or Goes in Circles","DART-VLN adds two lightweight test-time controls to vision-language navigation agents, cutting backtracking and stale memory reads without any retraining.","A new framework patches two of the most persistent failure modes in navigation AI — without touching the model's weights.\n\nResearchers introduced DART-VLN, a training-free control layer for discrete vision-language navigation agents. These are systems that follow natural-language instructions to move through environments, deciding step by step which way to go. The problem: even well-trained agents tend to get tripped up at test time by two specific bugs. First, they over-rely on old memory — stale observations from earlier in a route that no longer reflect where the agent actually is. Second, they backtrack inefficiently, reversing direction repeatedly when uncertain. DART-VLN applies Test-Time Memory Decay to down-weight outdated evidence at the memory read step, and Anti-Loop Regularization to penalize immediate direction reversals during action selection. Neither technique adds learnable parameters or alters the underlying model.\n\nThe results on standard benchmarks R2R and REVERIE show the combination produces shorter paths, lower runtime, and better navigation scores in key test conditions. That matters because the usual fix for a struggling AI system is more training data or a larger model — both expensive. A plug-in layer that improves reliability on frozen backbones is a cheaper path, and one that could extend the useful life of already-deployed systems.\n\nThe skeptic's note: benchmark gains on R2R and REVERIE are a familiar currency in navigation research, and real-world environments tend to punish assumptions those datasets quietly bake in — so how well this holds outside controlled settings remains an open question.","[\"ai\",\"robotics\",\"navigation\",\"research\"]","2026-07-02T04:00:00.000Z","2026-07-02T06:37:44.868Z","2026-07-02T06:37:47.877Z","published",null,[],"ai",[24,26,27,28],"robotics","navigation","research",[30],{"name":31,"url":32},"arXiv cs.AI","https:\u002F\u002Farxiv.org\u002Fabs\u002F2607.01043",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"]