[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"branding":3,"analytics":7,"article-ai-food-models-can-name-your-meal-but-not-count-its-calories":10,"sections":48},{"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":38,"tags":39,"sources":43,"feedback":47,"feedback_at":22,"cost_usd":47,"total_tokens":47},4561,"ai-food-models-can-name-your-meal-but-not-count-its-calories","AI Food Models Can Name Your Meal But Not Count Its Calories","A new benchmark finds vision-language models fail badly at nutritional reasoning and may give dangerous advice to diabetic users.","A new benchmark exposes a wide gap between what AI vision models see on a plate and what they actually understand about it.\n\nResearchers introduced OmniFood-Bench, built from a dataset of 100,000 food images, to test whether large vision-language models can do more than identify dishes. The benchmark runs models through three tiers: recognizing ingredients and cooking methods, estimating portion sizes and nutrients, and generating advice for users with specific health conditions. The results are blunt — models score near-human on naming food but collapse when asked to estimate mass or tailor advice for high-risk profiles like diabetics. The researchers call this the \"Semantic-Physical Gap,\" a mismatch between labeling what something is and reasoning about what it contains.\n\nThe gap matters because AI-assisted dietary tools are already reaching consumers, and the failure modes here are not benign. Hallucinating a safe recommendation for a diabetic user is not a classification error — it is a patient safety problem. Benchmarks that only test food recognition have been obscuring this risk.\n\nFood AI has long oversold its clinical readiness. The pattern of \"impressive demo, brittle edge case\" is familiar from medical imaging and radiology tools that performed well on curated test sets and poorly in deployment. OmniFood-Bench at least gives the field a harder target to aim at — though a benchmark only changes outcomes if model developers actually train against it.","[\"ai\",\"health\",\"benchmarks\",\"vision-language-models\"]","2026-07-10T04:00:00.000Z","2026-07-10T04:54:33.713Z","2026-07-10T04:54:36.636Z","published",null,[24,30,34],{"id":25,"reviewer":26,"round":27,"reason":28,"status":29},"editor-r1","editor",1,"Three unresolved problems: (1) the article calls an arXiv preprint 'peer-reviewed' — arXiv is a preprint server with no peer review, a factual error the writer introduced; (2) the model identifiers GPT-5.1 and Gemini-3 Flash match the pattern of unverifiable model names flagged in editorial policy and must be verified against known public releases before publication; (3) the paper's code repository is hosted on anonymous.4open.science, meaning authorship is anonymous, and the article neither fla","resolved",{"id":31,"reviewer":26,"round":32,"reason":33,"status":29},"editor-r2",2,"The draft correctly flags the peer-review status and the anonymous repository, and surfaces the unverifiable model identifiers as a caveat — but editorial policy requires those ambiguities to be resolved before submission, not merely disclosed in the body; kill the specific per-model framing entirely or hold the piece until the model names can be verified against known public releases.",{"id":35,"reviewer":26,"round":36,"reason":37,"status":29},"editor-r3",3,"The draft still names and tests specific models from the source (implicitly via 'six state-of-the-art models') while the source itself cites identifiers — gpt-5.1, gemini-3-flash, qwen3-vl-8B — that cannot be verified against any known public release; the draft must either omit per-model framing entirely or hold until model names are confirmed, per [editor-r2], which this draft has not resolved.","ai",[38,40,41,42],"health","benchmarks","vision-language-models",[44],{"name":45,"url":46},"arXiv cs.AI","https:\u002F\u002Farxiv.org\u002Fabs\u002F2607.08423",0,{"sections":49},[50,54,59,64,69,74,79,84,89,94,99,103,108,113],{"name":51,"slug":38,"count":52,"latest_published_at":53},"AI",2590,"2026-07-16T04:00:00.000Z",{"name":55,"slug":56,"count":57,"latest_published_at":58},"Security","security",294,"2026-07-15T19:59:48.000Z",{"name":60,"slug":61,"count":62,"latest_published_at":63},"Deals","deals",179,"2026-06-29T20:02:07.000Z",{"name":65,"slug":66,"count":67,"latest_published_at":68},"Policy","policy",158,"2026-07-16T00:02:48.000Z",{"name":70,"slug":71,"count":72,"latest_published_at":73},"Hardware","hardware",122,"2026-07-14T19:46:26.000Z",{"name":75,"slug":76,"count":77,"latest_published_at":78},"Consumer Tech","consumer-tech",93,"2026-07-13T13:20:48.000Z",{"name":80,"slug":81,"count":82,"latest_published_at":83},"Software","software",70,"2026-07-13T19:52:25.000Z",{"name":85,"slug":86,"count":87,"latest_published_at":88},"Science","science",66,"2026-07-10T10:29:37.000Z",{"name":90,"slug":91,"count":92,"latest_published_at":93},"Dev Tools","dev-tools",59,"2026-07-07T04:00:00.000Z",{"name":95,"slug":96,"count":97,"latest_published_at":98},"Gaming","gaming",41,"2026-07-09T04:00:00.000Z",{"name":100,"slug":101,"count":97,"latest_published_at":102},"Startups","startups","2026-06-29T20:55:50.000Z",{"name":104,"slug":105,"count":106,"latest_published_at":107},"General","general",29,"2026-07-10T22:28:58.000Z",{"name":109,"slug":110,"count":111,"latest_published_at":112},"Reviews","reviews",20,"2026-06-24T12:00:01.000Z",{"name":114,"slug":115,"count":116,"latest_published_at":117},"How-To","how-to",6,"2026-06-16T09:00:00.000Z"]