[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"branding":3,"analytics":7,"article-new-benchmark-pushes-ai-to-name-who-harmful-memes-target":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},5470,"new-benchmark-pushes-ai-to-name-who-harmful-memes-target","New Benchmark Pushes AI to Name Who Harmful Memes Target","A new benchmark and training method push AI models to correctly identify who and what a harmful meme attacks, not just flag it as harmful.","Researchers have built a system that doesn't just flag a meme as harmful - it identifies who the meme attacks and where the attack is hiding in the image or text.\n\nThe team behind HarmTrace built a new dataset, Meme3W, by combining and re-annotating several existing harmful-meme collections with human-verified labels. They also introduced a stricter evaluation metric, Joint Record Accuracy (JRA), which only counts a prediction as correct if the model gets the harmfulness call right and correctly names the target category, target entity, text mention, and visual region. Testing existing multimodal large language models on this metric exposed a big gap: models that could spot harmful content often still misidentified who or what was being attacked. HarmTrace closes that gap through a two-step training process, and on the Qwen3-VL-8B model, JRA jumped from 17.58% to 52.51%.\n\nThat gap matters more than a leaderboard number suggests. Most content-moderation systems today report a single accuracy figure for harmfulness, which hides whether the system actually understood the joke or just heard alarm bells and flagged it. A platform that can't say who a meme attacks can't reliably route it for human review, defend a takedown decision, or avoid overcorrecting.\n\nNone of this ships in a moderation tool yet - it's a research benchmark, and a near-tripling on one metric for one model backbone is not the same as a production-ready fix.","[\"ai\",\"research\",\"content-moderation\",\"harmful-memes\"]","2026-08-18T04:00:00.000Z","2026-08-18T21:35:31.517Z","2026-08-18T21:35:43.463Z","published",null,[],"ai",[24,26,27,28],"research","content-moderation","harmful-memes",[30],{"name":31,"url":32},"arXiv cs.AI","https:\u002F\u002Farxiv.org\u002Fabs\u002F2608.16622",0,{"sections":35},[36,40,44,49,54,59,64,69,74,78,83,88,93,98],{"name":37,"slug":24,"count":38,"latest_published_at":39},"AI",3293,"2026-08-20T04:00:00.000Z",{"name":41,"slug":42,"count":43,"latest_published_at":39},"Security","security",435,{"name":45,"slug":46,"count":47,"latest_published_at":48},"Policy","policy",210,"2026-08-19T09:32:27.000Z",{"name":50,"slug":51,"count":52,"latest_published_at":53},"Deals","deals",179,"2026-06-29T20:02:07.000Z",{"name":55,"slug":56,"count":57,"latest_published_at":58},"Hardware","hardware",140,"2026-08-19T18:25:42.000Z",{"name":60,"slug":61,"count":62,"latest_published_at":63},"Consumer Tech","consumer-tech",95,"2026-08-18T16:05:00.000Z",{"name":65,"slug":66,"count":67,"latest_published_at":68},"Science","science",90,"2026-08-19T18:41:02.000Z",{"name":70,"slug":71,"count":72,"latest_published_at":73},"Software","software",73,"2026-08-18T07:51:50.000Z",{"name":75,"slug":76,"count":77,"latest_published_at":18},"Dev Tools","dev-tools",69,{"name":79,"slug":80,"count":81,"latest_published_at":82},"Startups","startups",47,"2026-08-19T19:13:46.000Z",{"name":84,"slug":85,"count":86,"latest_published_at":87},"Gaming","gaming",41,"2026-07-09T04:00:00.000Z",{"name":89,"slug":90,"count":91,"latest_published_at":92},"General","general",33,"2026-08-18T22:18:13.000Z",{"name":94,"slug":95,"count":96,"latest_published_at":97},"Reviews","reviews",20,"2026-06-24T12:00:01.000Z",{"name":99,"slug":100,"count":101,"latest_published_at":102},"How-To","how-to",6,"2026-06-16T09:00:00.000Z"]