{"id":85468,"date":"2026-07-18T14:30:54","date_gmt":"2026-07-18T11:30:54","guid":{"rendered":"https:\/\/twodots.gr\/?p=85468"},"modified":"2026-07-18T14:30:56","modified_gmt":"2026-07-18T11:30:56","slug":"mag-neo-benchmark-ai-agents-sto-web","status":"publish","type":"post","link":"https:\/\/twodots.gr\/en\/mag-neo-benchmark-ai-agents-sto-web\/","title":{"rendered":"MAG: \u03a4\u03b9 \u03b1\u03c0\u03bf\u03ba\u03b1\u03bb\u03cd\u03c0\u03c4\u03b5\u03b9 \u03c4\u03bf \u03bd\u03ad\u03bf benchmark \u03b3\u03b9\u03b1 \u03c4\u03bf\u03c5\u03c2 AI agents \u03c3\u03c4\u03bf web"},"content":{"rendered":"<div class=\"td-article-toc\"><div class=\"td-toc-title\">Contents<\/div><ul><li><a href=\"#ti-metra-mag\">\u03a4\u03b9 \u03bc\u03b5\u03c4\u03c1\u03ac \u03c0\u03c1\u03b1\u03b3\u03bc\u03b1\u03c4\u03b9\u03ba\u03ac \u03c4\u03bf MAG<\/a><\/li><li><a href=\"#giati-screenshots-web-agents\">\u0393\u03b9\u03b1\u03c4\u03af \u03c4\u03b1 screenshots \u03b1\u03bb\u03bb\u03ac\u03b6\u03bf\u03c5\u03bd \u03c4\u03b7\u03bd \u03b1\u03be\u03b9\u03bf\u03bb\u03cc\u03b3\u03b7\u03c3\u03b7 \u03c4\u03c9\u03bd web agents<\/a><\/li><li><a href=\"#set-of-mark-i-syntetagmenes\">Set-of-Mark \u03ae \u03c3\u03c5\u03bd\u03c4\u03b5\u03c4\u03b1\u03b3\u03bc\u03ad\u03bd\u03b5\u03c2: \u03b4\u03cd\u03bf \u03b4\u03b9\u03b1\u03c6\u03bf\u03c1\u03b5\u03c4\u03b9\u03ba\u03bf\u03af \u03c4\u03c1\u03cc\u03c0\u03bf\u03b9 grounding<\/a><\/li><li><a href=\"#dataset-kai-elegchos\">\u03a0\u03ce\u03c2 \u03b4\u03b7\u03bc\u03b9\u03bf\u03c5\u03c1\u03b3\u03ae\u03b8\u03b7\u03ba\u03b5 \u03c4\u03bf benchmark \u03ba\u03b1\u03b9 \u03c4\u03b9 \u03b1\u03ba\u03c1\u03b9\u03b2\u03ce\u03c2 \u03c0\u03b5\u03c1\u03b9\u03bb\u03b1\u03bc\u03b2\u03ac\u03bd\u03b5\u03b9<\/a><\/li><li><a href=\"#apotelesmata-mag\">\u03a4\u03b9 \u03b1\u03c0\u03bf\u03ba\u03b1\u03bb\u03cd\u03c0\u03c4\u03bf\u03c5\u03bd \u03c4\u03b1 \u03b1\u03c0\u03bf\u03c4\u03b5\u03bb\u03ad\u03c3\u03bc\u03b1\u03c4\u03b1 \u03c4\u03bf\u03c5 MAG<\/a><\/li><li><a href=\"#makria-tasks-kindynos\">\u0393\u03b9\u03b1\u03c4\u03af \u03c4\u03b1 \u03bc\u03b5\u03b3\u03ac\u03bb\u03b1 web tasks \u03c0\u03b1\u03c1\u03b1\u03bc\u03ad\u03bd\u03bf\u03c5\u03bd \u03bf \u03b1\u03b4\u03cd\u03bd\u03b1\u03bc\u03bf\u03c2 \u03ba\u03c1\u03af\u03ba\u03bf\u03c2<\/a><\/li><li><a href=\"#epicheirimatiki-aksia\">\u03a4\u03b9 \u03c3\u03b7\u03bc\u03b1\u03af\u03bd\u03bf\u03c5\u03bd \u03c4\u03b1 \u03b5\u03c5\u03c1\u03ae\u03bc\u03b1\u03c4\u03b1 \u03b3\u03b9\u03b1 e-commerce \u03ba\u03b1\u03b9 \u03b5\u03c0\u03b9\u03c7\u03b5\u03b9\u03c1\u03b7\u03bc\u03b1\u03c4\u03b9\u03ba\u03ad\u03c2 \u03c1\u03bf\u03ad\u03c2<\/a><\/li><li><a href=\"#aksiologisi-dikou-agent\">\u03a0\u03ce\u03c2 \u03bd\u03b1 \u03b1\u03be\u03b9\u03bf\u03bb\u03bf\u03b3\u03ae\u03c3\u03b5\u03c4\u03b5 \u03c4\u03bf\u03bd \u03b4\u03b9\u03ba\u03cc \u03c3\u03b1\u03c2 AI agent<\/a><\/li><li><a href=\"#production-guardrails\">\u03a4\u03b1 guardrails \u03c0\u03c1\u03b9\u03bd \u03b1\u03c0\u03cc \u03c4\u03b7\u03bd \u03c0\u03b1\u03c1\u03b1\u03b3\u03c9\u03b3\u03b9\u03ba\u03ae \u03c7\u03c1\u03ae\u03c3\u03b7<\/a><\/li><li><a href=\"#mag-kritirio-agoras\">\u03a4\u03bf MAG \u03b1\u03bb\u03bb\u03ac\u03b6\u03b5\u03b9 \u03c4\u03bf \u03b2\u03b1\u03c3\u03b9\u03ba\u03cc \u03ba\u03c1\u03b9\u03c4\u03ae\u03c1\u03b9\u03bf \u03b1\u03b3\u03bf\u03c1\u03ac\u03c2 \u03b5\u03bd\u03cc\u03c2 AI agent<\/a><\/li><\/ul><\/div>\n\n\n<div class=\"td-article-lede\"><p><strong>\u03a3\u03cd\u03bd\u03c4\u03bf\u03bc\u03b7 \u03b1\u03c0\u03ac\u03bd\u03c4\u03b7\u03c3\u03b7:<\/strong> \u03c4\u03bf MAG \u03b5\u03af\u03bd\u03b1\u03b9 \u03ad\u03bd\u03b1 \u03bd\u03ad\u03bf \u03b5\u03c1\u03b5\u03c5\u03bd\u03b7\u03c4\u03b9\u03ba\u03cc benchmark \u03c0\u03bf\u03c5 \u03b6\u03b7\u03c4\u03ac \u03b1\u03c0\u03cc \u03ad\u03bd\u03b1\u03bd AI agent \u03bd\u03b1 \u03ba\u03ac\u03bd\u03b5\u03b9 \u03b4\u03cd\u03bf \u03c0\u03c1\u03ac\u03b3\u03bc\u03b1\u03c4\u03b1 \u03c4\u03b1\u03c5\u03c4\u03cc\u03c7\u03c1\u03bf\u03bd\u03b1: \u03bd\u03b1 \u03bf\u03bb\u03bf\u03ba\u03bb\u03b7\u03c1\u03ce\u03c3\u03b5\u03b9 \u03bc\u03b9\u03b1 \u03c0\u03bf\u03bb\u03c5\u03c3\u03c4\u03b1\u03b4\u03b9\u03b1\u03ba\u03ae \u03b5\u03c1\u03b3\u03b1\u03c3\u03af\u03b1 \u03c3\u03b5 \u03c0\u03c1\u03b1\u03b3\u03bc\u03b1\u03c4\u03b9\u03ba\u03cc web \u03c0\u03b5\u03c1\u03b9\u03b2\u03ac\u03bb\u03bb\u03bf\u03bd \u03ba\u03b1\u03b9 \u03bd\u03b1 \u03b3\u03c1\u03ac\u03c8\u03b5\u03b9, \u03c3\u03b5 \u03ba\u03ac\u03b8\u03b5 \u03b2\u03ae\u03bc\u03b1, \u03bc\u03b9\u03b1 \u03c3\u03cd\u03bd\u03c4\u03bf\u03bc\u03b7 \u03bf\u03b4\u03b7\u03b3\u03af\u03b1 \u03c0\u03bf\u03c5 \u03b8\u03b1 \u03bc\u03c0\u03bf\u03c1\u03bf\u03cd\u03c3\u03b5 \u03bd\u03b1 \u03b1\u03ba\u03bf\u03bb\u03bf\u03c5\u03b8\u03ae\u03c3\u03b5\u03b9 \u03ac\u03bd\u03b8\u03c1\u03c9\u03c0\u03bf\u03c2. \u03a4\u03bf \u03ba\u03b1\u03bb\u03cd\u03c4\u03b5\u03c1\u03bf \u03c3\u03cd\u03c3\u03c4\u03b7\u03bc\u03b1 \u03c4\u03b7\u03c2 \u03bc\u03b5\u03bb\u03ad\u03c4\u03b7\u03c2 \u03bf\u03bb\u03bf\u03ba\u03bb\u03ae\u03c1\u03c9\u03c3\u03b5 \u03c4\u03bf 37,4% \u03c4\u03c9\u03bd test tasks. \u0386\u03c1\u03b1 \u03bf\u03b9 web agents \u03ad\u03c7\u03bf\u03c5\u03bd \u03ba\u03ac\u03bd\u03b5\u03b9 \u03bf\u03c5\u03c3\u03b9\u03b1\u03c3\u03c4\u03b9\u03ba\u03ae \u03c0\u03c1\u03cc\u03bf\u03b4\u03bf, \u03b1\u03bb\u03bb\u03ac \u03b1\u03c0\u03ad\u03c7\u03bf\u03c5\u03bd \u03b1\u03ba\u03cc\u03bc\u03b7 \u03b1\u03c0\u03cc \u03c4\u03b7\u03bd \u03b1\u03bd\u03b5\u03be\u03ad\u03bb\u03b5\u03b3\u03ba\u03c4\u03b7 \u03b1\u03c5\u03c4\u03bf\u03bd\u03bf\u03bc\u03af\u03b1 \u03c0\u03bf\u03c5 \u03c5\u03c0\u03cc\u03c3\u03c7\u03bf\u03bd\u03c4\u03b1\u03b9 \u03c0\u03bf\u03bb\u03bb\u03ac demos. \u0393\u03b9\u03b1 \u03bc\u03b9\u03b1 \u03b5\u03c0\u03b9\u03c7\u03b5\u03af\u03c1\u03b7\u03c3\u03b7, \u03c4\u03bf \u03ba\u03c1\u03af\u03c3\u03b9\u03bc\u03bf \u03bc\u03ac\u03b8\u03b7\u03bc\u03b1 \u03b5\u03af\u03bd\u03b1\u03b9 \u03cc\u03c4\u03b9 \u03b7 \u03b1\u03be\u03b9\u03bf\u03bb\u03cc\u03b3\u03b7\u03c3\u03b7 \u03c0\u03c1\u03ad\u03c0\u03b5\u03b9 \u03bd\u03b1 \u03bc\u03b5\u03c4\u03c1\u03ac \u03b1\u03c0\u03bf\u03c4\u03ad\u03bb\u03b5\u03c3\u03bc\u03b1, \u03b1\u03ba\u03c1\u03af\u03b2\u03b5\u03b9\u03b1 \u03b5\u03bd\u03b5\u03c1\u03b3\u03b5\u03b9\u03ce\u03bd, \u03c0\u03bf\u03b9\u03cc\u03c4\u03b7\u03c4\u03b1 \u03bf\u03b4\u03b7\u03b3\u03b9\u03ce\u03bd, \u03bc\u03ae\u03ba\u03bf\u03c2 \u03c1\u03bf\u03ae\u03c2 \u03ba\u03b1\u03b9 \u03b1\u03c3\u03c6\u03b1\u03bb\u03ae \u03b1\u03c0\u03bf\u03c4\u03c5\u03c7\u03af\u03b1 \u03c0\u03ac\u03bd\u03c9 \u03c3\u03c4\u03b1 \u03b4\u03b9\u03ba\u03ac \u03c4\u03b7\u03c2 \u03c3\u03c5\u03c3\u03c4\u03ae\u03bc\u03b1\u03c4\u03b1.<\/p><\/div>\n\n<p>\u0388\u03bd\u03b1\u03c2 web agent \u03bc\u03c0\u03bf\u03c1\u03b5\u03af \u03bd\u03b1 \u03c6\u03b1\u03af\u03bd\u03b5\u03c4\u03b1\u03b9 \u03b5\u03bd\u03c4\u03c5\u03c0\u03c9\u03c3\u03b9\u03b1\u03ba\u03cc\u03c2 \u03cc\u03c4\u03b1\u03bd \u03b1\u03bd\u03bf\u03af\u03b3\u03b5\u03b9 \u03c3\u03b5\u03bb\u03af\u03b4\u03b5\u03c2, \u03c3\u03c5\u03bc\u03c0\u03bb\u03b7\u03c1\u03ce\u03bd\u03b5\u03b9 \u03c0\u03b5\u03b4\u03af\u03b1 \u03ba\u03b1\u03b9 \u03c0\u03b1\u03c4\u03ac \u03ba\u03bf\u03c5\u03bc\u03c0\u03b9\u03ac. \u0391\u03c5\u03c4\u03cc \u03cc\u03bc\u03c9\u03c2 \u03b4\u03b5\u03bd \u03b1\u03c1\u03ba\u03b5\u03af \u03b3\u03b9\u03b1 \u03bd\u03b1 \u03c7\u03c1\u03b7\u03c3\u03b9\u03bc\u03bf\u03c0\u03bf\u03b9\u03b7\u03b8\u03b5\u03af \u03c3\u03b5 \u03c0\u03b1\u03c1\u03b1\u03b3\u03b3\u03b5\u03bb\u03af\u03b5\u03c2, CRM, \u03ba\u03b1\u03bc\u03c0\u03ac\u03bd\u03b9\u03b5\u03c2, \u03b5\u03c0\u03b9\u03c3\u03c4\u03c1\u03bf\u03c6\u03ad\u03c2, \u03b1\u03bd\u03b1\u03c6\u03bf\u03c1\u03ad\u03c2 \u03ae \u03b5\u03c3\u03c9\u03c4\u03b5\u03c1\u03b9\u03ba\u03ad\u03c2 \u03b5\u03b3\u03ba\u03c1\u03af\u03c3\u03b5\u03b9\u03c2. \u0397 \u03b5\u03c0\u03b9\u03c7\u03b5\u03b9\u03c1\u03b7\u03bc\u03b1\u03c4\u03b9\u03ba\u03ae \u03b1\u03be\u03af\u03b1 \u03c0\u03c1\u03bf\u03ba\u03cd\u03c0\u03c4\u03b5\u03b9 \u03bc\u03cc\u03bd\u03bf \u03cc\u03c4\u03b1\u03bd \u03bf agent \u03c6\u03c4\u03ac\u03bd\u03b5\u03b9 \u03c3\u03c4\u03bf \u03c3\u03c9\u03c3\u03c4\u03cc \u03c4\u03b5\u03bb\u03b9\u03ba\u03cc \u03b1\u03c0\u03bf\u03c4\u03ad\u03bb\u03b5\u03c3\u03bc\u03b1, \u03b1\u03c6\u03ae\u03bd\u03b5\u03b9 \u03ba\u03b1\u03c4\u03b1\u03bd\u03bf\u03b7\u03c4\u03cc \u03af\u03c7\u03bd\u03bf\u03c2 \u03b5\u03bd\u03b5\u03c1\u03b3\u03b5\u03b9\u03ce\u03bd \u03ba\u03b1\u03b9 \u03c3\u03c4\u03b1\u03bc\u03b1\u03c4\u03ac \u03bc\u03b5 \u03b5\u03bb\u03b5\u03b3\u03c7\u03cc\u03bc\u03b5\u03bd\u03bf \u03c4\u03c1\u03cc\u03c0\u03bf \u03cc\u03c4\u03b1\u03bd \u03c4\u03bf \u03c0\u03b5\u03c1\u03b9\u03b2\u03ac\u03bb\u03bb\u03bf\u03bd \u03b4\u03b5\u03bd \u03c4\u03b1\u03b9\u03c1\u03b9\u03ac\u03b6\u03b5\u03b9 \u03c3\u03c4\u03b9\u03c2 \u03c0\u03c1\u03bf\u03c3\u03b4\u03bf\u03ba\u03af\u03b5\u03c2 \u03c4\u03bf\u03c5.<\/p>\n\n\n<div class=\"td-article-note\"><p><strong>\u03a0\u03ce\u03c2 \u03bd\u03b1 \u03b4\u03b9\u03b1\u03b2\u03ac\u03c3\u03b5\u03c4\u03b5 \u03c4\u03b1 \u03b5\u03c5\u03c1\u03ae\u03bc\u03b1\u03c4\u03b1:<\/strong> \u03c4\u03bf \u03ac\u03c1\u03b8\u03c1\u03bf \u03b2\u03b1\u03c3\u03af\u03b6\u03b5\u03c4\u03b1\u03b9 \u03c3\u03c4\u03b7\u03bd \u03ad\u03ba\u03b4\u03bf\u03c3\u03b7 arXiv v3 \u03c4\u03b7\u03c2 16\u03b7\u03c2 \u0399\u03bf\u03c5\u03bb\u03af\u03bf\u03c5 2026 \u03ba\u03b1\u03b9 \u03c3\u03c4\u03b1 \u03b1\u03c0\u03bf\u03c4\u03b5\u03bb\u03ad\u03c3\u03bc\u03b1\u03c4\u03b1 \u03c0\u03bf\u03c5 \u03b1\u03bd\u03b1\u03c6\u03ad\u03c1\u03bf\u03c5\u03bd \u03bf\u03b9 \u03c3\u03c5\u03b3\u03b3\u03c1\u03b1\u03c6\u03b5\u03af\u03c2. \u0394\u03b5\u03bd \u03c0\u03c1\u03cc\u03ba\u03b5\u03b9\u03c4\u03b1\u03b9 \u03b3\u03b9\u03b1 \u03b1\u03bd\u03b5\u03be\u03ac\u03c1\u03c4\u03b7\u03c4\u03b7 \u03b1\u03bd\u03b1\u03c0\u03b1\u03c1\u03b1\u03b3\u03c9\u03b3\u03ae \u03ae \u03c0\u03c1\u03cc\u03b2\u03bb\u03b5\u03c8\u03b7 ROI. \u0397 \u03c3\u03b5\u03bb\u03af\u03b4\u03b1 \u03c4\u03bf\u03c5 paper \u03b1\u03bd\u03b1\u03c6\u03ad\u03c1\u03b5\u03b9 \u03cc\u03c4\u03b9 benchmark, harness \u03ba\u03b1\u03b9 model checkpoints \u03c0\u03c1\u03cc\u03ba\u03b5\u03b9\u03c4\u03b1\u03b9 \u03bd\u03b1 \u03b4\u03b9\u03b1\u03c4\u03b5\u03b8\u03bf\u03cd\u03bd, \u03b5\u03c0\u03bf\u03bc\u03ad\u03bd\u03c9\u03c2 \u03c4\u03bf MAG \u03c0\u03c1\u03ad\u03c0\u03b5\u03b9 \u03c3\u03ae\u03bc\u03b5\u03c1\u03b1 \u03bd\u03b1 \u03b4\u03b9\u03b1\u03b2\u03ac\u03b6\u03b5\u03c4\u03b1\u03b9 \u03c9\u03c2 \u03b5\u03c1\u03b5\u03c5\u03bd\u03b7\u03c4\u03b9\u03ba\u03cc \u03c0\u03bb\u03b1\u03af\u03c3\u03b9\u03bf \u03b1\u03be\u03b9\u03bf\u03bb\u03cc\u03b3\u03b7\u03c3\u03b7\u03c2 \u03ba\u03b1\u03b9 \u03cc\u03c7\u03b9 \u03c9\u03c2 \u03ad\u03c4\u03bf\u03b9\u03bc\u03bf \u03b5\u03bc\u03c0\u03bf\u03c1\u03b9\u03ba\u03cc \u03b5\u03c1\u03b3\u03b1\u03bb\u03b5\u03af\u03bf.<\/p><\/div>\n\n\n<h2 id=\"ti-metra-mag\">\u03a4\u03b9 \u03bc\u03b5\u03c4\u03c1\u03ac \u03c0\u03c1\u03b1\u03b3\u03bc\u03b1\u03c4\u03b9\u03ba\u03ac \u03c4\u03bf MAG<\/h2>\n\n<p>\u03a4\u03b1 \u03b1\u03c1\u03c7\u03b9\u03ba\u03ac MAG \u03c3\u03b7\u03bc\u03b1\u03af\u03bd\u03bf\u03c5\u03bd Multimodal Action and Guide. \u03a3\u03b5 \u03ba\u03ac\u03b8\u03b5 \u03b2\u03ae\u03bc\u03b1 \u03bf agent \u03b2\u03bb\u03ad\u03c0\u03b5\u03b9 \u03c4\u03b7\u03bd \u03c4\u03c1\u03ad\u03c7\u03bf\u03c5\u03c3\u03b1 \u03c3\u03b5\u03bb\u03af\u03b4\u03b1 \u03c9\u03c2 screenshot, \u03bb\u03b1\u03bc\u03b2\u03ac\u03bd\u03b5\u03b9 \u03c4\u03bf \u03b9\u03c3\u03c4\u03bf\u03c1\u03b9\u03ba\u03cc \u03c4\u03c9\u03bd \u03bf\u03b4\u03b7\u03b3\u03b9\u03ce\u03bd \u03c0\u03bf\u03c5 \u03ad\u03c7\u03b5\u03b9 \u03ae\u03b4\u03b7 \u03b3\u03c1\u03ac\u03c8\u03b5\u03b9 \u03ba\u03b1\u03b9 \u03b5\u03c0\u03b9\u03c3\u03c4\u03c1\u03ad\u03c6\u03b5\u03b9 \u03ad\u03bd\u03b1 \u03b6\u03b5\u03cd\u03b3\u03bf\u03c2: \u03c4\u03b7\u03bd \u03b5\u03c0\u03cc\u03bc\u03b5\u03bd\u03b7 \u03b5\u03bd\u03ad\u03c1\u03b3\u03b5\u03b9\u03b1 \u03ba\u03b1\u03b9 \u03bc\u03b9\u03b1 \u03b1\u03bd\u03b8\u03c1\u03ce\u03c0\u03b9\u03bd\u03b7 \u03bf\u03b4\u03b7\u03b3\u03af\u03b1. \u0397 \u03b5\u03bd\u03ad\u03c1\u03b3\u03b5\u03b9\u03b1 \u03bc\u03c0\u03bf\u03c1\u03b5\u03af \u03bd\u03b1 \u03b5\u03af\u03bd\u03b1\u03b9 click, \u03c0\u03bb\u03b7\u03ba\u03c4\u03c1\u03bf\u03bb\u03cc\u03b3\u03b7\u03c3\u03b7, \u03b5\u03c0\u03b9\u03bb\u03bf\u03b3\u03ae, scroll, Enter, \u03b5\u03c0\u03b9\u03c3\u03c4\u03c1\u03bf\u03c6\u03ae \u03ae \u03c4\u03b5\u03c1\u03bc\u03b1\u03c4\u03b9\u03c3\u03bc\u03cc\u03c2. \u0397 \u03bf\u03b4\u03b7\u03b3\u03af\u03b1 \u03c0\u03c1\u03ad\u03c0\u03b5\u03b9 \u03bd\u03b1 \u03b5\u03be\u03b7\u03b3\u03b5\u03af \u03c4\u03b9 \u03ba\u03ac\u03bd\u03b5\u03b9 \u03bf \u03c7\u03c1\u03ae\u03c3\u03c4\u03b7\u03c2 \u03c3\u03c4\u03b7\u03bd \u03bf\u03c1\u03b1\u03c4\u03ae \u03c3\u03b5\u03bb\u03af\u03b4\u03b1 \u03c7\u03c9\u03c1\u03af\u03c2 \u03bd\u03b1 \u03b1\u03c0\u03bf\u03ba\u03b1\u03bb\u03cd\u03c0\u03c4\u03b5\u03b9 \u03c4\u03b5\u03c7\u03bd\u03b9\u03ba\u03ac identifiers, \u03b1\u03c1\u03b9\u03b8\u03bc\u03bf\u03cd\u03c2 marks \u03ae \u03c3\u03c5\u03bd\u03c4\u03b5\u03c4\u03b1\u03b3\u03bc\u03ad\u03bd\u03b5\u03c2 pixel.<\/p>\n\n<p>\u0397 \u03b4\u03b9\u03c0\u03bb\u03ae \u03b1\u03c0\u03b1\u03af\u03c4\u03b7\u03c3\u03b7 \u03b5\u03af\u03bd\u03b1\u03b9 \u03c3\u03b7\u03bc\u03b1\u03bd\u03c4\u03b9\u03ba\u03ae \u03b5\u03c0\u03b5\u03b9\u03b4\u03ae \u03b5\u03bd\u03ce\u03bd\u03b5\u03b9 \u03b4\u03cd\u03bf \u03c0\u03c1\u03bf\u03b2\u03bb\u03ae\u03bc\u03b1\u03c4\u03b1 \u03c0\u03bf\u03c5 \u03c3\u03c5\u03bd\u03ae\u03b8\u03c9\u03c2 \u03b1\u03be\u03b9\u03bf\u03bb\u03bf\u03b3\u03bf\u03cd\u03bd\u03c4\u03b1\u03b9 \u03c7\u03c9\u03c1\u03b9\u03c3\u03c4\u03ac. \u03a4\u03b1 \u03ba\u03bb\u03b1\u03c3\u03b9\u03ba\u03ac web-agent benchmarks \u03c1\u03c9\u03c4\u03bf\u03cd\u03bd \u03b1\u03bd \u03bf\u03bb\u03bf\u03ba\u03bb\u03b7\u03c1\u03ce\u03b8\u03b7\u03ba\u03b5 \u03c4\u03bf task. \u03a4\u03b1 \u03b5\u03c1\u03b3\u03b1\u03bb\u03b5\u03af\u03b1 digital adoption \u03ba\u03b1\u03b9 \u03bf\u03b9 \u03ad\u03c1\u03b5\u03c5\u03bd\u03b5\u03c2 \u03b3\u03b9\u03b1 in-app guidance \u03b5\u03be\u03b5\u03c4\u03ac\u03b6\u03bf\u03c5\u03bd \u03b1\u03bd \u03b4\u03b7\u03bc\u03b9\u03bf\u03c5\u03c1\u03b3\u03ae\u03b8\u03b7\u03ba\u03b5 \u03c7\u03c1\u03ae\u03c3\u03b9\u03bc\u03b7 \u03bf\u03b4\u03b7\u03b3\u03af\u03b1. \u03a4\u03bf MAG \u03b6\u03b7\u03c4\u03ac \u03b1\u03c0\u03cc \u03c4\u03bf \u03af\u03b4\u03b9\u03bf \u03c3\u03cd\u03c3\u03c4\u03b7\u03bc\u03b1 \u03bd\u03b1 \u03b5\u03bd\u03b5\u03c1\u03b3\u03ae\u03c3\u03b5\u03b9 \u03ba\u03b1\u03b9 \u03bd\u03b1 \u03b5\u03be\u03b7\u03b3\u03ae\u03c3\u03b5\u03b9, \u03ce\u03c3\u03c4\u03b5 \u03bc\u03b9\u03b1 \u03b5\u03c0\u03b9\u03c4\u03c5\u03c7\u03b7\u03bc\u03ad\u03bd\u03b7 \u03b4\u03b9\u03b1\u03b4\u03c1\u03bf\u03bc\u03ae \u03bd\u03b1 \u03bc\u03c0\u03bf\u03c1\u03b5\u03af \u03bd\u03b1 \u03bc\u03b5\u03c4\u03b1\u03c4\u03c1\u03b1\u03c0\u03b5\u03af \u03c3\u03b5 \u03b5\u03c0\u03b1\u03bd\u03b1\u03c7\u03c1\u03b7\u03c3\u03b9\u03bc\u03bf\u03c0\u03bf\u03b9\u03ae\u03c3\u03b9\u03bc\u03bf \u03bf\u03b4\u03b7\u03b3\u03cc \u03b3\u03b9\u03b1 \u03c4\u03bf\u03bd \u03b5\u03c0\u03cc\u03bc\u03b5\u03bd\u03bf \u03c7\u03c1\u03ae\u03c3\u03c4\u03b7.<\/p>\n\n<p>\u03a4\u03bf \u03b1\u03c0\u03bf\u03c4\u03ad\u03bb\u03b5\u03c3\u03bc\u03b1 \u03b4\u03b5\u03bd \u03ba\u03c1\u03af\u03bd\u03b5\u03c4\u03b1\u03b9 \u03b1\u03c0\u03cc \u03c4\u03bf \u03c0\u03cc\u03c3\u03bf \u03c0\u03b5\u03b9\u03c3\u03c4\u03b9\u03ba\u03ac \u03b3\u03c1\u03ac\u03c6\u03b5\u03b9 \u03c4\u03bf \u03bc\u03bf\u03bd\u03c4\u03ad\u03bb\u03bf. \u0397 \u03b5\u03c0\u03b9\u03c4\u03c5\u03c7\u03af\u03b1 \u03b5\u03bb\u03ad\u03b3\u03c7\u03b5\u03c4\u03b1\u03b9 \u03bb\u03b5\u03b9\u03c4\u03bf\u03c5\u03c1\u03b3\u03b9\u03ba\u03ac \u03c3\u03c4\u03b7\u03bd \u03c4\u03b5\u03bb\u03b9\u03ba\u03ae \u03ba\u03b1\u03c4\u03ac\u03c3\u03c4\u03b1\u03c3\u03b7 \u03c4\u03bf\u03c5 website. \u03a0\u03b1\u03c1\u03ac\u03bb\u03bb\u03b7\u03bb\u03b1, \u03bf\u03b9 \u03bf\u03b4\u03b7\u03b3\u03af\u03b5\u03c2 \u03c3\u03c5\u03b3\u03ba\u03c1\u03af\u03bd\u03bf\u03bd\u03c4\u03b1\u03b9 \u03bc\u03b5 \u03b1\u03bd\u03b8\u03c1\u03ce\u03c0\u03b9\u03bd\u03b1 \u03b5\u03c0\u03b1\u03bb\u03b7\u03b8\u03b5\u03c5\u03bc\u03ad\u03bd\u03b1 gold guides \u03ba\u03b1\u03b9 \u03c0\u03b5\u03c1\u03bd\u03bf\u03cd\u03bd format gate. \u0388\u03c4\u03c3\u03b9 \u03bc\u03b9\u03b1 \u03ac\u03c8\u03bf\u03b3\u03b7 \u03c0\u03b5\u03c1\u03b9\u03b3\u03c1\u03b1\u03c6\u03ae \u03b4\u03b5\u03bd \u03ba\u03b1\u03bb\u03cd\u03c0\u03c4\u03b5\u03b9 \u03bc\u03b9\u03b1 \u03b1\u03c0\u03bf\u03c4\u03c5\u03c7\u03b7\u03bc\u03ad\u03bd\u03b7 \u03b5\u03bd\u03ad\u03c1\u03b3\u03b5\u03b9\u03b1 \u03ba\u03b1\u03b9 \u03ad\u03bd\u03b1 \u03c3\u03c9\u03c3\u03c4\u03cc click \u03b4\u03b5\u03bd \u03b1\u03c1\u03ba\u03b5\u03af \u03b1\u03bd \u03b7 \u03c3\u03c5\u03bd\u03bf\u03b4\u03b5\u03c5\u03c4\u03b9\u03ba\u03ae \u03bf\u03b4\u03b7\u03b3\u03af\u03b1 \u03b5\u03af\u03bd\u03b1\u03b9 \u03b1\u03ba\u03b1\u03c4\u03b1\u03bd\u03cc\u03b7\u03c4\u03b7 \u03ae \u03b4\u03b9\u03b1\u03c1\u03c1\u03ad\u03b5\u03b9 \u03b5\u03c3\u03c9\u03c4\u03b5\u03c1\u03b9\u03ba\u03ac \u03c3\u03c4\u03bf\u03b9\u03c7\u03b5\u03af\u03b1.<\/p>\n\n\n<div class=\"td-decision-band\"><p class=\"td-decision-label\">Demo \u2260 \u03b5\u03c0\u03b1\u03bb\u03b7\u03b8\u03b5\u03c5\u03bc\u03ad\u03bd\u03b7 \u03b5\u03c0\u03b9\u03c7\u03b5\u03b9\u03c1\u03b7\u03bc\u03b1\u03c4\u03b9\u03ba\u03ae \u03c1\u03bf\u03ae<\/p><strong>\u0397 \u03c6\u03c5\u03c3\u03b9\u03ba\u03ae \u03ba\u03af\u03bd\u03b7\u03c3\u03b7 \u03c4\u03bf\u03c5 \u03c0\u03bf\u03bd\u03c4\u03b9\u03ba\u03b9\u03bf\u03cd \u03b4\u03b5\u03bd \u03b1\u03c0\u03bf\u03b4\u03b5\u03b9\u03ba\u03bd\u03cd\u03b5\u03b9 \u03cc\u03c4\u03b9 \u03b5\u03c0\u03b9\u03c4\u03b5\u03cd\u03c7\u03b8\u03b7\u03ba\u03b5 \u03bf \u03c3\u03c9\u03c3\u03c4\u03cc\u03c2 \u03c3\u03c4\u03cc\u03c7\u03bf\u03c2.<\/strong><p>\u0393\u03b9\u03b1 \u03bd\u03b1 \u03b1\u03be\u03b9\u03bf\u03bb\u03bf\u03b3\u03b7\u03b8\u03b5\u03af \u03ad\u03bd\u03b1\u03c2 web agent \u03c7\u03c1\u03b5\u03b9\u03ac\u03b6\u03bf\u03bd\u03c4\u03b1\u03b9 \u03b3\u03bd\u03c9\u03c3\u03c4\u03ae \u03b1\u03c1\u03c7\u03b9\u03ba\u03ae \u03ba\u03b1\u03c4\u03ac\u03c3\u03c4\u03b1\u03c3\u03b7, \u03b1\u03bd\u03c4\u03b9\u03ba\u03b5\u03b9\u03bc\u03b5\u03bd\u03b9\u03ba\u03cc\u03c2 \u03ad\u03bb\u03b5\u03b3\u03c7\u03bf\u03c2 \u03c4\u03bf\u03c5 \u03c4\u03b5\u03bb\u03b9\u03ba\u03bf\u03cd \u03b1\u03c0\u03bf\u03c4\u03b5\u03bb\u03ad\u03c3\u03bc\u03b1\u03c4\u03bf\u03c2, \u03ba\u03b1\u03c4\u03b1\u03b3\u03c1\u03b1\u03c6\u03ae \u03ba\u03ac\u03b8\u03b5 \u03b2\u03ae\u03bc\u03b1\u03c4\u03bf\u03c2 \u03ba\u03b1\u03b9 \u03c3\u03b1\u03c6\u03ad\u03c2 \u03ba\u03c1\u03b9\u03c4\u03ae\u03c1\u03b9\u03bf \u03c4\u03b5\u03c1\u03bc\u03b1\u03c4\u03b9\u03c3\u03bc\u03bf\u03cd. \u03a7\u03c9\u03c1\u03af\u03c2 \u03b1\u03c5\u03c4\u03ac, \u03ad\u03bd\u03b1 \u03b5\u03bd\u03c4\u03c5\u03c0\u03c9\u03c3\u03b9\u03b1\u03ba\u03cc walkthrough \u03bc\u03c0\u03bf\u03c1\u03b5\u03af \u03bd\u03b1 \u03ba\u03c1\u03cd\u03b2\u03b5\u03b9 \u03bb\u03ac\u03b8\u03bf\u03c2 \u03b4\u03b5\u03b4\u03bf\u03bc\u03ad\u03bd\u03b1, \u03c0\u03c1\u03cc\u03c9\u03c1\u03bf finish \u03ae \u03b5\u03bd\u03ad\u03c1\u03b3\u03b5\u03b9\u03b1 \u03c3\u03b5 \u03bb\u03ac\u03b8\u03bf\u03c2 \u03bb\u03bf\u03b3\u03b1\u03c1\u03b9\u03b1\u03c3\u03bc\u03cc.<\/p><\/div>\n\n\n<h2 id=\"giati-screenshots-web-agents\">\u0393\u03b9\u03b1\u03c4\u03af \u03c4\u03b1 screenshots \u03b1\u03bb\u03bb\u03ac\u03b6\u03bf\u03c5\u03bd \u03c4\u03b7\u03bd \u03b1\u03be\u03b9\u03bf\u03bb\u03cc\u03b3\u03b7\u03c3\u03b7 \u03c4\u03c9\u03bd web agents<\/h2>\n\n<p>\u03a0\u03bf\u03bb\u03bb\u03ac \u03c3\u03c5\u03c3\u03c4\u03ae\u03bc\u03b1\u03c4\u03b1 web automation \u03b4\u03b9\u03b1\u03b2\u03ac\u03b6\u03bf\u03c5\u03bd \u03c4\u03bf DOM \u03ae \u03ad\u03bd\u03b1 accessibility tree. \u0391\u03c5\u03c4\u03ad\u03c2 \u03bf\u03b9 \u03b1\u03bd\u03b1\u03c0\u03b1\u03c1\u03b1\u03c3\u03c4\u03ac\u03c3\u03b5\u03b9\u03c2 \u03c0\u03c1\u03bf\u03c3\u03c6\u03ad\u03c1\u03bf\u03c5\u03bd \u03ba\u03b1\u03b8\u03b1\u03c1\u03ae \u03b4\u03bf\u03bc\u03ae \u03ba\u03b1\u03b9 \u03c3\u03c5\u03c7\u03bd\u03ac \u03ba\u03ac\u03bd\u03bf\u03c5\u03bd \u03b5\u03c5\u03ba\u03bf\u03bb\u03cc\u03c4\u03b5\u03c1\u03bf \u03c4\u03bf\u03bd \u03b5\u03bd\u03c4\u03bf\u03c0\u03b9\u03c3\u03bc\u03cc \u03b5\u03bd\u03cc\u03c2 \u03ba\u03bf\u03c5\u03bc\u03c0\u03b9\u03bf\u03cd \u03ae \u03c0\u03b5\u03b4\u03af\u03bf\u03c5. \u039f \u03ac\u03bd\u03b8\u03c1\u03c9\u03c0\u03bf\u03c2 \u03cc\u03bc\u03c9\u03c2 \u03b4\u03b5\u03bd \u03c7\u03c1\u03b7\u03c3\u03b9\u03bc\u03bf\u03c0\u03bf\u03b9\u03b5\u03af \u03c4\u03bf website \u03c9\u03c2 \u03bb\u03af\u03c3\u03c4\u03b1 \u03b1\u03c0\u03cc nodes. \u0392\u03bb\u03ad\u03c0\u03b5\u03b9 \u03c4\u03b7 rendered \u03c3\u03b5\u03bb\u03af\u03b4\u03b1, \u03c4\u03b7 \u03b8\u03ad\u03c3\u03b7 \u03c4\u03c9\u03bd \u03c3\u03c4\u03bf\u03b9\u03c7\u03b5\u03af\u03c9\u03bd, \u03c4\u03b7\u03bd \u03bf\u03c0\u03c4\u03b9\u03ba\u03ae \u03b9\u03b5\u03c1\u03b1\u03c1\u03c7\u03af\u03b1, \u03c4\u03b1 overlays, \u03c4\u03b1 \u03bc\u03b5\u03bd\u03bf\u03cd \u03c0\u03bf\u03c5 \u03b1\u03bd\u03bf\u03af\u03b3\u03bf\u03c5\u03bd, \u03c4\u03b9\u03c2 \u03b1\u03bb\u03bb\u03b1\u03b3\u03ad\u03c2 \u03ba\u03b1\u03c4\u03ac\u03c3\u03c4\u03b1\u03c3\u03b7\u03c2 \u03ba\u03b1\u03b9 \u03c4\u03bf \u03c0\u03b5\u03c1\u03b9\u03b5\u03c7\u03cc\u03bc\u03b5\u03bd\u03bf \u03c0\u03bf\u03c5 \u03b5\u03bc\u03c6\u03b1\u03bd\u03af\u03b6\u03b5\u03c4\u03b1\u03b9 \u03bc\u03b5\u03c4\u03ac \u03b1\u03c0\u03cc scroll.<\/p>\n\n<p>\u03a4\u03bf MAG \u03ba\u03c1\u03b1\u03c4\u03ac \u03c4\u03bf screenshot \u03c3\u03c4\u03bf \u03ba\u03ad\u03bd\u03c4\u03c1\u03bf \u03c4\u03b7\u03c2 \u03c0\u03b1\u03c1\u03b1\u03c4\u03ae\u03c1\u03b7\u03c3\u03b7\u03c2. \u0397 \u03b5\u03c0\u03b9\u03bb\u03bf\u03b3\u03ae \u03b1\u03c5\u03c4\u03ae \u03c6\u03ad\u03c1\u03bd\u03b5\u03b9 \u03c4\u03b7\u03bd \u03b1\u03be\u03b9\u03bf\u03bb\u03cc\u03b3\u03b7\u03c3\u03b7 \u03c0\u03b9\u03bf \u03ba\u03bf\u03bd\u03c4\u03ac \u03c3\u03b5 \u03cc,\u03c4\u03b9 \u03c3\u03c5\u03bc\u03b2\u03b1\u03af\u03bd\u03b5\u03b9 \u03c3\u03b5 \u03ad\u03bd\u03b1 \u03c0\u03c1\u03b1\u03b3\u03bc\u03b1\u03c4\u03b9\u03ba\u03cc e-shop, CRM \u03ae help desk, \u03cc\u03c0\u03bf\u03c5 pop-ups, responsive layouts, sticky headers \u03ba\u03b1\u03b9 \u03b4\u03c5\u03bd\u03b1\u03bc\u03b9\u03ba\u03ac panels \u03bc\u03c0\u03bf\u03c1\u03bf\u03cd\u03bd \u03bd\u03b1 \u03b1\u03bb\u03bb\u03ac\u03be\u03bf\u03c5\u03bd \u03c4\u03b7 \u03b4\u03b9\u03b1\u03b8\u03ad\u03c3\u03b9\u03bc\u03b7 \u03b5\u03bd\u03ad\u03c1\u03b3\u03b5\u03b9\u03b1. \u0388\u03bd\u03b1\u03c2 agent \u03c0\u03bf\u03c5 \u03b2\u03b1\u03c3\u03af\u03b6\u03b5\u03c4\u03b1\u03b9 \u03bc\u03cc\u03bd\u03bf \u03c3\u03b5 \u03ba\u03b5\u03b9\u03bc\u03b5\u03bd\u03b9\u03ba\u03ae \u03b4\u03bf\u03bc\u03ae \u03bc\u03c0\u03bf\u03c1\u03b5\u03af \u03bd\u03b1 \u03b5\u03bd\u03c4\u03bf\u03c0\u03af\u03c3\u03b5\u03b9 \u03c3\u03c9\u03c3\u03c4\u03cc element \u03b1\u03bb\u03bb\u03ac \u03bd\u03b1 \u03bc\u03b7\u03bd \u03b1\u03bd\u03c4\u03b9\u03bb\u03b7\u03c6\u03b8\u03b5\u03af \u03cc\u03c4\u03b9 \u03b5\u03af\u03bd\u03b1\u03b9 \u03ba\u03c1\u03c5\u03bc\u03bc\u03ad\u03bd\u03bf, \u03ba\u03b1\u03bb\u03c5\u03bc\u03bc\u03ad\u03bd\u03bf \u03ae \u03bf\u03c0\u03c4\u03b9\u03ba\u03ac \u03b4\u03b9\u03b1\u03c6\u03bf\u03c1\u03b5\u03c4\u03b9\u03ba\u03cc \u03b1\u03c0\u03cc \u03b1\u03c5\u03c4\u03cc \u03c0\u03bf\u03c5 \u03c0\u03b5\u03c1\u03b9\u03bc\u03ad\u03bd\u03b5\u03b9 \u03bf \u03c7\u03c1\u03ae\u03c3\u03c4\u03b7\u03c2.<\/p>\n\n<p>\u0397 \u03bf\u03c0\u03c4\u03b9\u03ba\u03ae \u03b1\u03be\u03b9\u03bf\u03bb\u03cc\u03b3\u03b7\u03c3\u03b7 \u03b4\u03b5\u03bd \u03ba\u03b1\u03c4\u03b1\u03c1\u03b3\u03b5\u03af \u03c4\u03b7\u03bd \u03b1\u03be\u03af\u03b1 \u03c4\u03bf\u03c5 DOM. \u0394\u03b5\u03af\u03c7\u03bd\u03b5\u03b9 \u03cc\u03bc\u03c9\u03c2 \u03cc\u03c4\u03b9 \u03b7 \u03b1\u03be\u03b9\u03bf\u03c0\u03b9\u03c3\u03c4\u03af\u03b1 \u03b5\u03be\u03b1\u03c1\u03c4\u03ac\u03c4\u03b1\u03b9 \u03b1\u03c0\u03cc \u03c4\u03bf \u03ba\u03b1\u03bd\u03ac\u03bb\u03b9 \u03c0\u03b1\u03c1\u03b1\u03c4\u03ae\u03c1\u03b7\u03c3\u03b7\u03c2. \u03a3\u03b5 \u03c0\u03b1\u03c1\u03b1\u03b3\u03c9\u03b3\u03b9\u03ba\u03cc \u03c3\u03cd\u03c3\u03c4\u03b7\u03bc\u03b1 \u03bc\u03c0\u03bf\u03c1\u03b5\u03af \u03bd\u03b1 \u03c7\u03c1\u03b5\u03b9\u03ac\u03b6\u03b5\u03c4\u03b1\u03b9 \u03c3\u03c5\u03bd\u03b4\u03c5\u03b1\u03c3\u03bc\u03cc\u03c2 screenshot, \u03b4\u03bf\u03bc\u03b7\u03bc\u03ad\u03bd\u03c9\u03bd \u03b4\u03b5\u03b4\u03bf\u03bc\u03ad\u03bd\u03c9\u03bd, API \u03ba\u03b1\u03b9 deterministic checks. \u03a4\u03bf \u03ba\u03b1\u03c4\u03ac\u03bb\u03bb\u03b7\u03bb\u03bf \u03bc\u03b5\u03af\u03b3\u03bc\u03b1 \u03c0\u03c1\u03ad\u03c0\u03b5\u03b9 \u03bd\u03b1 \u03b4\u03bf\u03ba\u03b9\u03bc\u03ac\u03b6\u03b5\u03c4\u03b1\u03b9 \u03c3\u03c4\u03b9\u03c2 \u03af\u03b4\u03b9\u03b5\u03c2 \u03c3\u03b5\u03bb\u03af\u03b4\u03b5\u03c2 \u03ba\u03b1\u03b9 \u03c3\u03c4\u03b9\u03c2 \u03af\u03b4\u03b9\u03b5\u03c2 \u03b5\u03c1\u03b3\u03b1\u03c3\u03af\u03b5\u03c2 \u03c0\u03bf\u03c5 \u03b8\u03b1 \u03b1\u03bd\u03c4\u03b9\u03bc\u03b5\u03c4\u03c9\u03c0\u03af\u03c3\u03b5\u03b9 \u03bf agent, \u03cc\u03c7\u03b9 \u03bc\u03cc\u03bd\u03bf \u03c3\u03b5 \u03ad\u03bd\u03b1 \u03b3\u03b5\u03bd\u03b9\u03ba\u03cc benchmark.<\/p>\n\n<h2 id=\"set-of-mark-i-syntetagmenes\">Set-of-Mark \u03ae \u03c3\u03c5\u03bd\u03c4\u03b5\u03c4\u03b1\u03b3\u03bc\u03ad\u03bd\u03b5\u03c2: \u03b4\u03cd\u03bf \u03b4\u03b9\u03b1\u03c6\u03bf\u03c1\u03b5\u03c4\u03b9\u03ba\u03bf\u03af \u03c4\u03c1\u03cc\u03c0\u03bf\u03b9 grounding<\/h2>\n\n<p>\u03a4\u03bf MAG \u03c3\u03c5\u03b3\u03ba\u03c1\u03af\u03bd\u03b5\u03b9 \u03b4\u03cd\u03bf \u03c4\u03c1\u03cc\u03c0\u03bf\u03c5\u03c2 \u03bc\u03b5 \u03c4\u03bf\u03c5\u03c2 \u03bf\u03c0\u03bf\u03af\u03bf\u03c5\u03c2 \u03bf agent \u03c3\u03c5\u03bd\u03b4\u03ad\u03b5\u03b9 \u03c4\u03b7\u03bd \u03c0\u03c1\u03cc\u03b8\u03b5\u03c3\u03ae \u03c4\u03bf\u03c5 \u03bc\u03b5 \u03ad\u03bd\u03b1 \u03bf\u03c1\u03b1\u03c4\u03cc \u03c3\u03c4\u03bf\u03b9\u03c7\u03b5\u03af\u03bf. \u03a3\u03c4\u03bf Set-of-Mark, \u03b1\u03c1\u03b9\u03b8\u03bc\u03b7\u03bc\u03ad\u03bd\u03b5\u03c2 \u03c3\u03b7\u03bc\u03ac\u03bd\u03c3\u03b5\u03b9\u03c2 \u03c4\u03bf\u03c0\u03bf\u03b8\u03b5\u03c4\u03bf\u03cd\u03bd\u03c4\u03b1\u03b9 \u03c0\u03ac\u03bd\u03c9 \u03c3\u03c4\u03b1 \u03b4\u03b9\u03b1\u03b4\u03c1\u03b1\u03c3\u03c4\u03b9\u03ba\u03ac \u03c3\u03c4\u03bf\u03b9\u03c7\u03b5\u03af\u03b1 \u03ba\u03b1\u03b9 \u03bf agent \u03b5\u03c0\u03b9\u03bb\u03ad\u03b3\u03b5\u03b9 \u03c4\u03bf \u03ba\u03b1\u03c4\u03ac\u03bb\u03bb\u03b7\u03bb\u03bf mark. \u03a3\u03c4\u03b7 \u03bc\u03ad\u03b8\u03bf\u03b4\u03bf \u03c4\u03c9\u03bd \u03c3\u03c5\u03bd\u03c4\u03b5\u03c4\u03b1\u03b3\u03bc\u03ad\u03bd\u03c9\u03bd, \u03bf agent \u03b5\u03c0\u03b9\u03c3\u03c4\u03c1\u03ad\u03c6\u03b5\u03b9 \u03b1\u03c0\u03b5\u03c5\u03b8\u03b5\u03af\u03b1\u03c2 \u03c4\u03b7 \u03b8\u03ad\u03c3\u03b7 pixel \u03c3\u03c4\u03b7\u03bd \u03bf\u03c0\u03bf\u03af\u03b1 \u03c0\u03c1\u03ad\u03c0\u03b5\u03b9 \u03bd\u03b1 \u03b3\u03af\u03bd\u03b5\u03b9 \u03b7 \u03b5\u03bd\u03ad\u03c1\u03b3\u03b5\u03b9\u03b1. \u039a\u03b1\u03b9 \u03c3\u03c4\u03b9\u03c2 \u03b4\u03cd\u03bf \u03c0\u03b5\u03c1\u03b9\u03c0\u03c4\u03ce\u03c3\u03b5\u03b9\u03c2 \u03c4\u03bf screenshot, \u03c4\u03bf prompt, \u03c4\u03bf action budget \u03ba\u03b1\u03b9 \u03b7 \u03b2\u03b1\u03b8\u03bc\u03bf\u03bb\u03cc\u03b3\u03b7\u03c3\u03b7 \u03c0\u03b1\u03c1\u03b1\u03bc\u03ad\u03bd\u03bf\u03c5\u03bd \u03af\u03b4\u03b9\u03b1.<\/p>\n\n\n<div class=\"td-comparison\"><p class=\"td-comparison-title\">\u0394\u03cd\u03bf \u03c4\u03c1\u03cc\u03c0\u03bf\u03b9 \u03c3\u03c4\u03cc\u03c7\u03b5\u03c5\u03c3\u03b7\u03c2 \u03c4\u03b7\u03c2 \u03af\u03b4\u03b9\u03b1\u03c2 web \u03b5\u03bd\u03ad\u03c1\u03b3\u03b5\u03b9\u03b1\u03c2<\/p><div class=\"td-comparison-cards td-comparison-cards--horizontal\"><div class=\"td-comparison-grid td-comparison-grid--two\"><div class=\"td-platform-card\"><h3>Set-of-Mark<\/h3><p>\u039f agent \u03b5\u03c0\u03b9\u03bb\u03ad\u03b3\u03b5\u03b9 \u03bc\u03af\u03b1 \u03b1\u03c1\u03b9\u03b8\u03bc\u03b7\u03bc\u03ad\u03bd\u03b7 \u03c3\u03ae\u03bc\u03b1\u03bd\u03c3\u03b7 \u03c0\u03bf\u03c5 \u03b1\u03bd\u03c4\u03b9\u03c3\u03c4\u03bf\u03b9\u03c7\u03b5\u03af \u03c3\u03b5 \u03b4\u03b9\u03b1\u03b4\u03c1\u03b1\u03c3\u03c4\u03b9\u03ba\u03cc \u03c3\u03c4\u03bf\u03b9\u03c7\u03b5\u03af\u03bf. \u03a4\u03bf \u03b5\u03bd\u03b4\u03b9\u03ac\u03bc\u03b5\u03c3\u03bf layer \u03b1\u03c0\u03bb\u03bf\u03c0\u03bf\u03b9\u03b5\u03af \u03c4\u03bf grounding, \u03b1\u03bb\u03bb\u03ac \u03b5\u03be\u03b1\u03c1\u03c4\u03ac\u03c4\u03b1\u03b9 \u03b1\u03c0\u03cc \u03c4\u03bf \u03b1\u03bd \u03c4\u03b1 \u03c3\u03c9\u03c3\u03c4\u03ac \u03c3\u03c4\u03bf\u03b9\u03c7\u03b5\u03af\u03b1 \u03ad\u03c7\u03bf\u03c5\u03bd \u03b5\u03bd\u03c4\u03bf\u03c0\u03b9\u03c3\u03c4\u03b5\u03af \u03ba\u03b1\u03b9 \u03c3\u03b7\u03bc\u03b1\u03bd\u03b8\u03b5\u03af.<\/p><div class=\"td-badge-row\"><span class=\"td-badge\">Candidate elements<\/span><span class=\"td-badge\">\u0395\u03bb\u03b5\u03b3\u03c7\u03cc\u03bc\u03b5\u03bd\u03b7 \u03b5\u03c0\u03b9\u03bb\u03bf\u03b3\u03ae<\/span><\/div><\/div><div class=\"td-platform-card td-platform-card--navy\"><h3>\u03a3\u03c5\u03bd\u03c4\u03b5\u03c4\u03b1\u03b3\u03bc\u03ad\u03bd\u03b5\u03c2 pixel<\/h3><p>\u039f agent \u03c3\u03c4\u03bf\u03c7\u03b5\u03cd\u03b5\u03b9 \u03b1\u03c0\u03b5\u03c5\u03b8\u03b5\u03af\u03b1\u03c2 \u03c4\u03b7 \u03b8\u03ad\u03c3\u03b7 \u03c3\u03c4\u03b7\u03bd \u03bf\u03b8\u03cc\u03bd\u03b7. \u039c\u03b5\u03b9\u03ce\u03bd\u03b5\u03b9 \u03c4\u03b7\u03bd \u03b5\u03be\u03ac\u03c1\u03c4\u03b7\u03c3\u03b7 \u03b1\u03c0\u03cc \u03ad\u03bd\u03b1 marking layer, \u03b1\u03bb\u03bb\u03ac \u03b1\u03c0\u03b1\u03b9\u03c4\u03b5\u03af \u03b1\u03ba\u03c1\u03b9\u03b2\u03ae \u03bf\u03c0\u03c4\u03b9\u03ba\u03ae \u03c3\u03c4\u03cc\u03c7\u03b5\u03c5\u03c3\u03b7 \u03ba\u03b1\u03b9 \u03bc\u03c0\u03bf\u03c1\u03b5\u03af \u03bd\u03b1 \u03b5\u03c0\u03b7\u03c1\u03b5\u03b1\u03c3\u03c4\u03b5\u03af \u03b1\u03c0\u03cc viewport, scroll \u03ba\u03b1\u03b9 \u03b1\u03bb\u03bb\u03b1\u03b3\u03ad\u03c2 layout.<\/p><div class=\"td-badge-row\"><span class=\"td-badge\">Raw pixels<\/span><span class=\"td-badge\">\u0386\u03bc\u03b5\u03c3\u03bf grounding<\/span><\/div><\/div><\/div><\/div><\/div>\n\n\n<p>\u03a4\u03b1 \u03b1\u03c0\u03bf\u03c4\u03b5\u03bb\u03ad\u03c3\u03bc\u03b1\u03c4\u03b1 \u03b4\u03b5\u03bd \u03b1\u03bd\u03b1\u03b4\u03b5\u03b9\u03ba\u03bd\u03cd\u03bf\u03c5\u03bd \u03ad\u03bd\u03b1\u03bd \u03ba\u03b1\u03b8\u03bf\u03bb\u03b9\u03ba\u03cc \u03bd\u03b9\u03ba\u03b7\u03c4\u03ae. \u03a4\u03bf Gemini \u03b5\u03bc\u03c6\u03b1\u03bd\u03af\u03b6\u03b5\u03b9 \u03c3\u03b1\u03c6\u03ae \u03c0\u03c1\u03bf\u03c4\u03af\u03bc\u03b7\u03c3\u03b7 \u03c3\u03c4\u03bf Set-of-Mark, \u03b5\u03bd\u03ce \u03c4\u03b1 GPT-5.5 \u03ba\u03b1\u03b9 Claude \u03b4\u03b5\u03bd \u03c0\u03b1\u03c1\u03bf\u03c5\u03c3\u03b9\u03ac\u03b6\u03bf\u03c5\u03bd \u03b1\u03bd\u03c4\u03af\u03c3\u03c4\u03bf\u03b9\u03c7\u03b7 \u03c3\u03b7\u03bc\u03b1\u03bd\u03c4\u03b9\u03ba\u03ae \u03c0\u03c1\u03bf\u03c4\u03af\u03bc\u03b7\u03c3\u03b7. \u03a4\u03bf \u03b5\u03ba\u03c0\u03b1\u03b9\u03b4\u03b5\u03c5\u03bc\u03ad\u03bd\u03bf \u03bc\u03bf\u03bd\u03c4\u03ad\u03bb\u03bf 9B \u03b2\u03b5\u03bb\u03c4\u03b9\u03ce\u03bd\u03b5\u03c4\u03b1\u03b9 \u03ba\u03c5\u03c1\u03af\u03c9\u03c2 \u03bc\u03b5 Set-of-Mark. \u0393\u03b9\u03b1 \u03bc\u03b9\u03b1 \u03b5\u03c0\u03b9\u03c7\u03b5\u03af\u03c1\u03b7\u03c3\u03b7 \u03b1\u03c5\u03c4\u03cc \u03c3\u03b7\u03bc\u03b1\u03af\u03bd\u03b5\u03b9 \u03cc\u03c4\u03b9 \u03c4\u03bf grounding \u03b5\u03af\u03bd\u03b1\u03b9 \u03b1\u03c1\u03c7\u03b9\u03c4\u03b5\u03ba\u03c4\u03bf\u03bd\u03b9\u03ba\u03ae \u03b5\u03c0\u03b9\u03bb\u03bf\u03b3\u03ae \u03c0\u03bf\u03c5 \u03c0\u03c1\u03ad\u03c0\u03b5\u03b9 \u03bd\u03b1 \u03bc\u03b5\u03c4\u03c1\u03b7\u03b8\u03b5\u03af \u03b1\u03bd\u03ac \u03bc\u03bf\u03bd\u03c4\u03ad\u03bb\u03bf, browser, \u03b1\u03bd\u03ac\u03bb\u03c5\u03c3\u03b7, \u03c3\u03b5\u03bb\u03af\u03b4\u03b1 \u03ba\u03b1\u03b9 task.<\/p>\n\n<h2 id=\"dataset-kai-elegchos\">\u03a0\u03ce\u03c2 \u03b4\u03b7\u03bc\u03b9\u03bf\u03c5\u03c1\u03b3\u03ae\u03b8\u03b7\u03ba\u03b5 \u03c4\u03bf benchmark \u03ba\u03b1\u03b9 \u03c4\u03b9 \u03b1\u03ba\u03c1\u03b9\u03b2\u03ce\u03c2 \u03c0\u03b5\u03c1\u03b9\u03bb\u03b1\u03bc\u03b2\u03ac\u03bd\u03b5\u03b9<\/h2>\n\n<p>\u0397 \u03b2\u03ac\u03c3\u03b7 \u03c4\u03bf\u03c5 MAG \u03c0\u03c1\u03bf\u03ad\u03c1\u03c7\u03b5\u03c4\u03b1\u03b9 \u03b1\u03c0\u03cc 581 \u03b5\u03c0\u03b9\u03c4\u03c5\u03c7\u03b7\u03bc\u03ad\u03bd\u03b5\u03c2 demonstrations \u03c4\u03bf\u03c5 OpAgent \u03c0\u03ac\u03bd\u03c9 \u03c3\u03b5 \u03b5\u03c1\u03b3\u03b1\u03c3\u03af\u03b5\u03c2 \u03c4\u03bf\u03c5 WebArena. \u039f\u03b9 \u03b4\u03b9\u03b1\u03b4\u03c1\u03bf\u03bc\u03ad\u03c2 \u03b1\u03bd\u03b1\u03c0\u03b1\u03c1\u03ac\u03c7\u03b8\u03b7\u03ba\u03b1\u03bd \u03c3\u03b5 live sites, \u03ba\u03b1\u03c4\u03b1\u03b3\u03c1\u03ac\u03c6\u03b7\u03ba\u03b1\u03bd screenshots \u03ba\u03b1\u03b9 \u03b1\u03bd\u03c4\u03b9\u03c3\u03c4\u03bf\u03b9\u03c7\u03af\u03c3\u03c4\u03b7\u03ba\u03b1\u03bd \u03bf\u03b9 \u03b5\u03bd\u03ad\u03c1\u03b3\u03b5\u03b9\u03b5\u03c2 \u03c4\u03cc\u03c3\u03bf \u03c3\u03b5 Set-of-Mark \u03cc\u03c3\u03bf \u03ba\u03b1\u03b9 \u03c3\u03b5 \u03c3\u03c5\u03bd\u03c4\u03b5\u03c4\u03b1\u03b3\u03bc\u03ad\u03bd\u03b5\u03c2. \u039f\u03b9 \u03c3\u03c5\u03b3\u03b3\u03c1\u03b1\u03c6\u03b5\u03af\u03c2 \u03b1\u03bd\u03b1\u03c6\u03ad\u03c1\u03bf\u03c5\u03bd \u03cc\u03c4\u03b9 563 tasks \u03b5\u03c0\u03ad\u03b6\u03b7\u03c3\u03b1\u03bd \u03c4\u03b7\u03c2 \u03b4\u03b9\u03b1\u03b4\u03b9\u03ba\u03b1\u03c3\u03af\u03b1\u03c2 \u03b5\u03c0\u03b1\u03bd\u03b1\u03c3\u03c7\u03bf\u03bb\u03b9\u03b1\u03c3\u03bc\u03bf\u03cd \u03ba\u03b1\u03b9 \u03b1\u03c0\u03ad\u03ba\u03c4\u03b7\u03c3\u03b1\u03bd \u03c0\u03bb\u03ae\u03c1\u03b5\u03b9\u03c2 gold trajectories.<\/p>\n\n<p>\u03a4\u03bf \u03c4\u03b5\u03bb\u03b9\u03ba\u03cc \u03c5\u03bb\u03b9\u03ba\u03cc \u03c0\u03b5\u03c1\u03b9\u03bb\u03b1\u03bc\u03b2\u03ac\u03bd\u03b5\u03b9 4.760 \u03b2\u03ae\u03bc\u03b1\u03c4\u03b1 Set-of-Mark \u03ba\u03b1\u03b9 5.779 \u03b2\u03ae\u03bc\u03b1\u03c4\u03b1 \u03c3\u03c5\u03bd\u03c4\u03b5\u03c4\u03b1\u03b3\u03bc\u03ad\u03bd\u03c9\u03bd. \u0393\u03b9\u03b1 \u03ba\u03ac\u03b8\u03b5 \u03b2\u03ae\u03bc\u03b1 \u03b4\u03b7\u03bc\u03b9\u03bf\u03c5\u03c1\u03b3\u03ae\u03b8\u03b7\u03ba\u03b5 \u03bf\u03b4\u03b7\u03b3\u03af\u03b1 \u03bc\u03b5 \u03b2\u03bf\u03ae\u03b8\u03b5\u03b9\u03b1 \u03b3\u03bb\u03c9\u03c3\u03c3\u03b9\u03ba\u03bf\u03cd \u03bc\u03bf\u03bd\u03c4\u03ad\u03bb\u03bf\u03c5 \u03ba\u03b1\u03b9 \u03b1\u03ba\u03bf\u03bb\u03bf\u03cd\u03b8\u03b7\u03c3\u03b5 \u03b1\u03bd\u03b8\u03c1\u03ce\u03c0\u03b9\u03bd\u03b7 \u03b4\u03b9\u03cc\u03c1\u03b8\u03c9\u03c3\u03b7. \u0397 \u03b1\u03c1\u03c7\u03b9\u03ba\u03ae \u03ba\u03b1\u03c4\u03b1\u03bd\u03bf\u03bc\u03ae \u03c4\u03c9\u03bd 581 tasks \u03ae\u03c4\u03b1\u03bd 407 \u03b3\u03b9\u03b1 training \u03ba\u03b1\u03b9 174 \u03b3\u03b9\u03b1 test, \u03b5\u03bd\u03ce \u03bc\u03b5\u03c4\u03ac \u03c4\u03b1 \u03c6\u03af\u03bb\u03c4\u03c1\u03b1 \u03c4\u03b1 \u03c0\u03bb\u03ae\u03c1\u03c9\u03c2 \u03c3\u03c7\u03bf\u03bb\u03b9\u03b1\u03c3\u03bc\u03ad\u03bd\u03b1 tasks \u03ad\u03b3\u03b9\u03bd\u03b1\u03bd 396 \u03ba\u03b1\u03b9 167 \u03b1\u03bd\u03c4\u03af\u03c3\u03c4\u03bf\u03b9\u03c7\u03b1. \u0397 \u03b4\u03b9\u03ac\u03ba\u03c1\u03b9\u03c3\u03b7 \u03b5\u03af\u03bd\u03b1\u03b9 \u03c3\u03b7\u03bc\u03b1\u03bd\u03c4\u03b9\u03ba\u03ae \u03b3\u03b9\u03b1 \u03bd\u03b1 \u03bc\u03b7 \u03c3\u03c5\u03b3\u03c7\u03ad\u03bf\u03bd\u03c4\u03b1\u03b9 \u03c4\u03bf source split \u03bc\u03b5 \u03c4\u03bf \u03c4\u03b5\u03bb\u03b9\u03ba\u03cc annotated corpus.<\/p>\n\n<p>\u0397 \u03bc\u03b5\u03b8\u03bf\u03b4\u03bf\u03bb\u03bf\u03b3\u03af\u03b1 \u03b4\u03af\u03bd\u03b5\u03b9 \u03ad\u03bd\u03b1 \u03c7\u03c1\u03ae\u03c3\u03b9\u03bc\u03bf \u03c0\u03c1\u03cc\u03c4\u03c5\u03c0\u03bf \u03b3\u03b9\u03b1 \u03b5\u03c4\u03b1\u03b9\u03c1\u03b9\u03ba\u03ac tests: replay \u03c3\u03b5 \u03b5\u03bb\u03b5\u03b3\u03c7\u03cc\u03bc\u03b5\u03bd\u03bf \u03c0\u03b5\u03c1\u03b9\u03b2\u03ac\u03bb\u03bb\u03bf\u03bd, \u03b1\u03bd\u03c4\u03b9\u03ba\u03b5\u03b9\u03bc\u03b5\u03bd\u03b9\u03ba\u03cc\u03c2 checker, \u03ba\u03b1\u03c4\u03b1\u03b3\u03c1\u03b1\u03c6\u03ae screenshot, \u03c6\u03b9\u03bb\u03c4\u03c1\u03ac\u03c1\u03b9\u03c3\u03bc\u03b1 \u03b1\u03c3\u03c5\u03bd\u03b5\u03c0\u03b5\u03b9\u03ce\u03bd \u03ba\u03b1\u03b9 \u03b1\u03bd\u03b8\u03c1\u03ce\u03c0\u03b9\u03bd\u03b7 \u03b5\u03c0\u03b1\u03bb\u03ae\u03b8\u03b5\u03c5\u03c3\u03b7. \u0394\u03b5\u03bd \u03b1\u03c1\u03ba\u03b5\u03af \u03bd\u03b1 \u03c0\u03b1\u03c1\u03ac\u03b3\u03b5\u03b9 \u03ad\u03bd\u03b1 LLM test cases \u03ae \u03bf\u03b4\u03b7\u03b3\u03af\u03b5\u03c2 \u03ba\u03b1\u03b9 \u03bd\u03b1 \u03b8\u03b5\u03c9\u03c1\u03bf\u03cd\u03bd\u03c4\u03b1\u03b9 \u03b1\u03c5\u03c4\u03bf\u03bc\u03ac\u03c4\u03c9\u03c2 \u03c3\u03c9\u03c3\u03c4\u03ac. \u0397 \u03b1\u03bd\u03b8\u03c1\u03ce\u03c0\u03b9\u03bd\u03b7 \u03b1\u03bd\u03b1\u03c3\u03ba\u03cc\u03c0\u03b7\u03c3\u03b7 \u03c0\u03c1\u03ad\u03c0\u03b5\u03b9 \u03bd\u03b1 \u03b5\u03c3\u03c4\u03b9\u03ac\u03b6\u03b5\u03b9 \u03c4\u03cc\u03c3\u03bf \u03c3\u03c4\u03bf \u03b1\u03c0\u03bf\u03c4\u03ad\u03bb\u03b5\u03c3\u03bc\u03b1 \u03cc\u03c3\u03bf \u03ba\u03b1\u03b9 \u03c3\u03c4\u03bf \u03b1\u03bd \u03b7 \u03bf\u03b4\u03b7\u03b3\u03af\u03b1 \u03b1\u03bd\u03c4\u03b9\u03c3\u03c4\u03bf\u03b9\u03c7\u03b5\u03af \u03c3\u03c4\u03b7\u03bd \u03ba\u03b1\u03c4\u03ac\u03c3\u03c4\u03b1\u03c3\u03b7 \u03c0\u03bf\u03c5 \u03ad\u03b2\u03bb\u03b5\u03c0\u03b5 \u03c0\u03c1\u03b1\u03b3\u03bc\u03b1\u03c4\u03b9\u03ba\u03ac \u03bf \u03c7\u03c1\u03ae\u03c3\u03c4\u03b7\u03c2.<\/p>\n\n<h2 id=\"apotelesmata-mag\">\u03a4\u03b9 \u03b1\u03c0\u03bf\u03ba\u03b1\u03bb\u03cd\u03c0\u03c4\u03bf\u03c5\u03bd \u03c4\u03b1 \u03b1\u03c0\u03bf\u03c4\u03b5\u03bb\u03ad\u03c3\u03bc\u03b1\u03c4\u03b1 \u03c4\u03bf\u03c5 MAG<\/h2>\n\n<p>\u0397 \u03b9\u03c3\u03c7\u03c5\u03c1\u03cc\u03c4\u03b5\u03c1\u03b7 \u03b4\u03b9\u03b1\u03bc\u03cc\u03c1\u03c6\u03c9\u03c3\u03b7 \u03bf\u03bb\u03bf\u03ba\u03bb\u03ae\u03c1\u03c9\u03c3\u03b5 65 \u03b1\u03c0\u03cc \u03c4\u03b1 174 test tasks, \u03b4\u03b7\u03bb\u03b1\u03b4\u03ae 37,4%. \u03a4\u03bf \u03c0\u03bf\u03c3\u03bf\u03c3\u03c4\u03cc \u03b4\u03b5\u03af\u03c7\u03bd\u03b5\u03b9 \u03cc\u03c4\u03b9 \u03c4\u03b1 \u03ba\u03bf\u03c1\u03c5\u03c6\u03b1\u03af\u03b1 multimodal \u03bc\u03bf\u03bd\u03c4\u03ad\u03bb\u03b1 \u03bc\u03c0\u03bf\u03c1\u03bf\u03cd\u03bd \u03bd\u03b1 \u03c7\u03b5\u03b9\u03c1\u03b9\u03c3\u03c4\u03bf\u03cd\u03bd \u03ad\u03bd\u03b1 \u03bf\u03c5\u03c3\u03b9\u03b1\u03c3\u03c4\u03b9\u03ba\u03cc \u03bc\u03ad\u03c1\u03bf\u03c2 \u03c3\u03cd\u03bd\u03b8\u03b5\u03c4\u03c9\u03bd web \u03b5\u03c1\u03b3\u03b1\u03c3\u03b9\u03ce\u03bd. \u03a4\u03b1\u03c5\u03c4\u03cc\u03c7\u03c1\u03bf\u03bd\u03b1, \u03c0\u03b5\u03c1\u03b9\u03c3\u03c3\u03cc\u03c4\u03b5\u03c1\u03b1 \u03b1\u03c0\u03cc \u03ad\u03be\u03b9 \u03c3\u03c4\u03b1 \u03b4\u03ad\u03ba\u03b1 tasks \u03b1\u03c0\u03ad\u03c4\u03c5\u03c7\u03b1\u03bd \u03b1\u03ba\u03cc\u03bc\u03b7 \u03ba\u03b1\u03b9 \u03c3\u03c4\u03b7\u03bd \u03ba\u03b1\u03bb\u03cd\u03c4\u03b5\u03c1\u03b7 \u03c0\u03b5\u03c1\u03af\u03c0\u03c4\u03c9\u03c3\u03b7. \u0397 \u03b1\u03c0\u03cc\u03c3\u03c4\u03b1\u03c3\u03b7 \u03b1\u03bd\u03ac\u03bc\u03b5\u03c3\u03b1 \u03c3\u03b5 \u03ad\u03bd\u03b1 \u03b5\u03c0\u03b9\u03c4\u03c5\u03c7\u03b7\u03bc\u03ad\u03bd\u03bf demo \u03ba\u03b1\u03b9 \u03c3\u03b5 \u03c3\u03c4\u03b1\u03b8\u03b5\u03c1\u03ae \u03c0\u03b1\u03c1\u03b1\u03b3\u03c9\u03b3\u03b9\u03ba\u03ae \u03bb\u03b5\u03b9\u03c4\u03bf\u03c5\u03c1\u03b3\u03af\u03b1 \u03c0\u03b1\u03c1\u03b1\u03bc\u03ad\u03bd\u03b5\u03b9 \u03bc\u03b5\u03b3\u03ac\u03bb\u03b7.<\/p>\n\n\n<div class=\"td-chart td-chart--metrics\"><p class=\"td-chart-title\">\u03a4\u03b1 \u03b2\u03b1\u03c3\u03b9\u03ba\u03ac author-reported \u03bc\u03b5\u03b3\u03ad\u03b8\u03b7 \u03c4\u03bf\u03c5 MAG<\/p><p class=\"td-chart-intro\">\u039f\u03b9 \u03c4\u03b9\u03bc\u03ad\u03c2 \u03b1\u03c6\u03bf\u03c1\u03bf\u03cd\u03bd \u03c4\u03b7\u03bd \u03ad\u03ba\u03b4\u03bf\u03c3\u03b7 arXiv v3 \u03ba\u03b1\u03b9 \u03c0\u03c1\u03ad\u03c0\u03b5\u03b9 \u03bd\u03b1 \u03b4\u03b9\u03b1\u03b2\u03ac\u03b6\u03bf\u03bd\u03c4\u03b1\u03b9 \u03bc\u03b1\u03b6\u03af \u03bc\u03b5 \u03c4\u03bf \u03c3\u03c5\u03b3\u03ba\u03b5\u03ba\u03c1\u03b9\u03bc\u03ad\u03bd\u03bf task set, \u03c4\u03bf grounding \u03ba\u03b1\u03b9 \u03c4\u03bf \u03bc\u03bf\u03bd\u03c4\u03ad\u03bb\u03bf.<\/p><div class=\"td-metric-grid\"><div class=\"td-metric-card\"><div class=\"td-metric-value\">581<\/div><div class=\"td-metric-label\">source tasks<\/div><p>\u0395\u03c0\u03b1\u03bb\u03b7\u03b8\u03b5\u03c5\u03bc\u03ad\u03bd\u03b5\u03c2 demonstrations \u03b1\u03c0\u03cc \u03c4\u03bf WebArena \u03b1\u03c0\u03bf\u03c4\u03ad\u03bb\u03b5\u03c3\u03b1\u03bd \u03c4\u03b7\u03bd \u03b1\u03c1\u03c7\u03b9\u03ba\u03ae \u03b2\u03ac\u03c3\u03b7 \u03c4\u03bf\u03c5 benchmark.<\/p><\/div><div class=\"td-metric-card\"><div class=\"td-metric-value\">563<\/div><div class=\"td-metric-label\">annotated tasks<\/div><p>Tasks \u03b5\u03c0\u03ad\u03b6\u03b7\u03c3\u03b1\u03bd \u03c4\u03bf\u03c5 replay \u03ba\u03b1\u03b9 \u03c4\u03bf\u03c5 \u03b5\u03c0\u03b1\u03bd\u03b1\u03c3\u03c7\u03bf\u03bb\u03b9\u03b1\u03c3\u03bc\u03bf\u03cd \u03bc\u03b5 gold trajectories \u03ba\u03b1\u03b9 \u03bf\u03b4\u03b7\u03b3\u03af\u03b5\u03c2.<\/p><\/div><div class=\"td-metric-card\"><div class=\"td-metric-value\">37,4%<\/div><div class=\"td-metric-label\">\u03ba\u03b1\u03bb\u03cd\u03c4\u03b5\u03c1\u03bf success rate<\/div><p>\u0397 \u03b9\u03c3\u03c7\u03c5\u03c1\u03cc\u03c4\u03b5\u03c1\u03b7 \u03b4\u03b9\u03b1\u03bc\u03cc\u03c1\u03c6\u03c9\u03c3\u03b7 \u03bf\u03bb\u03bf\u03ba\u03bb\u03ae\u03c1\u03c9\u03c3\u03b5 65 \u03b1\u03c0\u03cc \u03c4\u03b1 174 test tasks.<\/p><\/div><div class=\"td-metric-card\"><div class=\"td-metric-value\">13,2%<\/div><div class=\"td-metric-label\">\u03b5\u03ba\u03c0\u03b1\u03b9\u03b4\u03b5\u03c5\u03bc\u03ad\u03bd\u03bf\u03c2 9B agent<\/div><p>\u03a4\u03bf expert-augmented GRPO \u03c3\u03c7\u03b5\u03b4\u03cc\u03bd \u03b4\u03b9\u03c0\u03bb\u03b1\u03c3\u03af\u03b1\u03c3\u03b5 \u03c4\u03bf Set-of-Mark success \u03b1\u03c0\u03cc 6,9%.<\/p><\/div><\/div><\/div>\n\n\n<p>\u03a4\u03bf supervised fine-tuning \u03ad\u03bc\u03b1\u03b8\u03b5 \u03c3\u03c4\u03bf \u03bc\u03bf\u03bd\u03c4\u03ad\u03bb\u03bf 9B \u03bd\u03b1 \u03b1\u03ba\u03bf\u03bb\u03bf\u03c5\u03b8\u03b5\u03af \u03c0\u03bf\u03bb\u03cd \u03ba\u03b1\u03bb\u03cd\u03c4\u03b5\u03c1\u03b1 \u03c4\u03bf output contract, \u03b1\u03bb\u03bb\u03ac \u03b4\u03b5\u03bd \u03c4\u03bf\u03c5 \u03ad\u03b4\u03c9\u03c3\u03b5 \u03b1\u03c0\u03cc \u03bc\u03cc\u03bd\u03bf \u03c4\u03bf\u03c5 \u03b5\u03c0\u03b1\u03c1\u03ba\u03ae \u03b9\u03ba\u03b1\u03bd\u03cc\u03c4\u03b7\u03c4\u03b1 \u03bf\u03bb\u03bf\u03ba\u03bb\u03ae\u03c1\u03c9\u03c3\u03b7\u03c2. \u03a3\u03c4\u03bf Set-of-Mark \u03b7 \u03b5\u03c0\u03b9\u03c4\u03c5\u03c7\u03af\u03b1 \u03ad\u03c6\u03c4\u03b1\u03c3\u03b5 \u03bc\u03cc\u03bb\u03b9\u03c2 \u03c4\u03bf 6,9%, \u03c7\u03b1\u03bc\u03b7\u03bb\u03cc\u03c4\u03b5\u03c1\u03b1 \u03b1\u03c0\u03cc \u03c4\u03bf 8,1% \u03c4\u03bf\u03c5 untuned base, \u03b5\u03c0\u03b5\u03b9\u03b4\u03ae \u03c4\u03bf \u03bc\u03bf\u03bd\u03c4\u03ad\u03bb\u03bf \u03ad\u03bc\u03b1\u03b8\u03b5 \u03c3\u03c5\u03c7\u03bd\u03ac \u03bd\u03b1 \u03b4\u03b7\u03bb\u03ce\u03bd\u03b5\u03b9 \u03c0\u03c1\u03cc\u03c9\u03c1\u03b1 finish. \u03a0\u03c1\u03cc\u03ba\u03b5\u03b9\u03c4\u03b1\u03b9 \u03b3\u03b9\u03b1 \u03c7\u03b1\u03c1\u03b1\u03ba\u03c4\u03b7\u03c1\u03b9\u03c3\u03c4\u03b9\u03ba\u03cc \u03c0\u03b1\u03c1\u03ac\u03b4\u03b5\u03b9\u03b3\u03bc\u03b1 \u03cc\u03c0\u03bf\u03c5 \u03b7 \u03c3\u03c5\u03bc\u03bc\u03cc\u03c1\u03c6\u03c9\u03c3\u03b7 \u03c3\u03c4\u03b7 \u03bc\u03bf\u03c1\u03c6\u03ae \u03b4\u03b5\u03bd \u03b9\u03c3\u03bf\u03b4\u03c5\u03bd\u03b1\u03bc\u03b5\u03af \u03bc\u03b5 \u03bb\u03b5\u03b9\u03c4\u03bf\u03c5\u03c1\u03b3\u03b9\u03ba\u03ae \u03b9\u03ba\u03b1\u03bd\u03cc\u03c4\u03b7\u03c4\u03b1.<\/p>\n\n<p>\u0397 \u03b1\u03c0\u03bb\u03ae GRPO \u03b5\u03ba\u03c0\u03b1\u03af\u03b4\u03b5\u03c5\u03c3\u03b7 \u03b5\u03c0\u03af\u03c3\u03b7\u03c2 \u03c3\u03c4\u03b1\u03bc\u03ac\u03c4\u03b7\u03c3\u03b5, \u03b5\u03c0\u03b5\u03b9\u03b4\u03ae \u03bf\u03bc\u03ac\u03b4\u03b5\u03c2 \u03c0\u03bf\u03c5 \u03b1\u03c0\u03bf\u03c4\u03cd\u03b3\u03c7\u03b1\u03bd\u03b1\u03bd \u03cc\u03bb\u03b5\u03c2 \u03b4\u03b5\u03bd \u03b4\u03b7\u03bc\u03b9\u03bf\u03c5\u03c1\u03b3\u03bf\u03cd\u03c3\u03b1\u03bd \u03c7\u03c1\u03ae\u03c3\u03b9\u03bc\u03b7 \u03b4\u03b9\u03b1\u03ba\u03cd\u03bc\u03b1\u03bd\u03c3\u03b7 reward. \u039c\u03b5 \u03c4\u03b7\u03bd \u03c0\u03c1\u03bf\u03c3\u03b8\u03ae\u03ba\u03b7 \u03b5\u03c0\u03b1\u03bb\u03b7\u03b8\u03b5\u03c5\u03bc\u03ad\u03bd\u03c9\u03bd expert trajectories, \u03c4\u03bf Set-of-Mark success \u03c4\u03bf\u03c5 9B agent \u03b1\u03bd\u03ad\u03b2\u03b7\u03ba\u03b5 \u03b1\u03c0\u03cc 6,9% \u03c3\u03b5 13,2% \u03ba\u03b1\u03b9 \u03b2\u03b5\u03bb\u03c4\u03b9\u03ce\u03b8\u03b7\u03ba\u03b5 \u03c0\u03b1\u03c1\u03ac\u03bb\u03bb\u03b7\u03bb\u03b1 \u03b7 \u03c0\u03bf\u03b9\u03cc\u03c4\u03b7\u03c4\u03b1 \u03c4\u03c9\u03bd \u03bf\u03b4\u03b7\u03b3\u03b9\u03ce\u03bd. \u0397 \u03c0\u03c1\u03cc\u03bf\u03b4\u03bf\u03c2 \u03b5\u03af\u03bd\u03b1\u03b9 \u03bf\u03c5\u03c3\u03b9\u03b1\u03c3\u03c4\u03b9\u03ba\u03ae \u03b3\u03b9\u03b1 \u03c4\u03b7\u03bd \u03ad\u03c1\u03b5\u03c5\u03bd\u03b1, \u03b1\u03bb\u03bb\u03ac \u03c4\u03bf \u03c4\u03b5\u03bb\u03b9\u03ba\u03cc \u03b5\u03c0\u03af\u03c0\u03b5\u03b4\u03bf \u03c0\u03b1\u03c1\u03b1\u03bc\u03ad\u03bd\u03b5\u03b9 \u03c0\u03bf\u03bb\u03cd \u03c7\u03b1\u03bc\u03b7\u03bb\u03cc \u03b3\u03b9\u03b1 \u03b1\u03bd\u03b5\u03be\u03ad\u03bb\u03b5\u03b3\u03ba\u03c4\u03b7 \u03c0\u03b1\u03c1\u03b1\u03b3\u03c9\u03b3\u03b9\u03ba\u03ae \u03c7\u03c1\u03ae\u03c3\u03b7.<\/p>\n\n<h2 id=\"makria-tasks-kindynos\">\u0393\u03b9\u03b1\u03c4\u03af \u03c4\u03b1 \u03bc\u03b5\u03b3\u03ac\u03bb\u03b1 web tasks \u03c0\u03b1\u03c1\u03b1\u03bc\u03ad\u03bd\u03bf\u03c5\u03bd \u03bf \u03b1\u03b4\u03cd\u03bd\u03b1\u03bc\u03bf\u03c2 \u03ba\u03c1\u03af\u03ba\u03bf\u03c2<\/h2>\n\n<p>\u039a\u03ac\u03b8\u03b5 \u03b5\u03c0\u03b9\u03c0\u03bb\u03ad\u03bf\u03bd \u03b2\u03ae\u03bc\u03b1 \u03c0\u03c1\u03bf\u03c3\u03b8\u03ad\u03c4\u03b5\u03b9 \u03c0\u03b9\u03b8\u03b1\u03bd\u03cc\u03c4\u03b7\u03c4\u03b1 \u03bb\u03ac\u03b8\u03bf\u03c5\u03c2. \u039f agent \u03bc\u03c0\u03bf\u03c1\u03b5\u03af \u03bd\u03b1 \u03c0\u03b1\u03c4\u03ae\u03c3\u03b5\u03b9 \u03bb\u03ac\u03b8\u03bf\u03c2 element, \u03bd\u03b1 \u03c7\u03ac\u03c3\u03b5\u03b9 \u03bc\u03b9\u03b1 \u03b1\u03bb\u03bb\u03b1\u03b3\u03ae \u03ba\u03b1\u03c4\u03ac\u03c3\u03c4\u03b1\u03c3\u03b7\u03c2, \u03bd\u03b1 \u03c7\u03c1\u03b7\u03c3\u03b9\u03bc\u03bf\u03c0\u03bf\u03b9\u03ae\u03c3\u03b5\u03b9 \u03c0\u03b1\u03bb\u03b9\u03cc context, \u03bd\u03b1 \u03b3\u03c1\u03ac\u03c8\u03b5\u03b9 \u03bf\u03b4\u03b7\u03b3\u03af\u03b1 \u03c0\u03bf\u03c5 \u03b4\u03b5\u03bd \u03c4\u03b1\u03b9\u03c1\u03b9\u03ac\u03b6\u03b5\u03b9 \u03c3\u03c4\u03b7 \u03bd\u03ad\u03b1 \u03bf\u03b8\u03cc\u03bd\u03b7 \u03ae \u03bd\u03b1 \u03c4\u03b5\u03c1\u03bc\u03b1\u03c4\u03af\u03c3\u03b5\u03b9 \u03c0\u03c1\u03cc\u03c9\u03c1\u03b1. \u03a3\u03c4\u03bf MAG, \u03cc\u03bb\u03b5\u03c2 \u03bf\u03b9 \u03c0\u03b1\u03c1\u03b1\u03bb\u03bb\u03b1\u03b3\u03ad\u03c2 \u03c4\u03bf\u03c5 \u03bc\u03bf\u03bd\u03c4\u03ad\u03bb\u03bf\u03c5 9B \u03bc\u03ad\u03bd\u03bf\u03c5\u03bd \u03ba\u03bf\u03bd\u03c4\u03ac \u03c3\u03c4\u03bf \u03bc\u03b7\u03b4\u03ad\u03bd \u03b3\u03b9\u03b1 \u03b5\u03c1\u03b3\u03b1\u03c3\u03af\u03b5\u03c2 \u03bc\u03b5 \u03c0\u03b5\u03c1\u03b9\u03c3\u03c3\u03cc\u03c4\u03b5\u03c1\u03b1 \u03b1\u03c0\u03cc \u03b5\u03bd\u03bd\u03ad\u03b1 gold steps. \u03a4\u03b1 \u03b9\u03c3\u03c7\u03c5\u03c1\u03cc\u03c4\u03b5\u03c1\u03b1 API models \u03b4\u03b9\u03b1\u03c4\u03b7\u03c1\u03bf\u03cd\u03bd \u03ba\u03b1\u03bb\u03cd\u03c4\u03b5\u03c1\u03b7 \u03b5\u03c0\u03af\u03b4\u03bf\u03c3\u03b7, \u03b1\u03bb\u03bb\u03ac \u03b1\u03ba\u03cc\u03bc\u03b7 \u03ba\u03b1\u03b9 \u03b5\u03ba\u03b5\u03af \u03ba\u03b9\u03bd\u03bf\u03cd\u03bd\u03c4\u03b1\u03b9 \u03c0\u03b5\u03c1\u03af\u03c0\u03bf\u03c5 \u03c3\u03c4\u03bf 26%-34%.<\/p>\n\n<p>\u03a4\u03bf \u03bc\u03ae\u03ba\u03bf\u03c2 \u03b4\u03b5\u03bd \u03b5\u03af\u03bd\u03b1\u03b9 \u03b1\u03c0\u03bb\u03ce\u03c2 \u03c4\u03b5\u03c7\u03bd\u03b9\u03ba\u03cc \u03c3\u03c4\u03b1\u03c4\u03b9\u03c3\u03c4\u03b9\u03ba\u03cc. \u03a3\u03b5 \u03bc\u03b9\u03b1 \u03b5\u03c0\u03b9\u03c7\u03b5\u03af\u03c1\u03b7\u03c3\u03b7 \u03b1\u03bd\u03c4\u03b9\u03c3\u03c4\u03bf\u03b9\u03c7\u03b5\u03af \u03c3\u03b5 \u03b1\u03bb\u03c5\u03c3\u03af\u03b4\u03b5\u03c2 \u03cc\u03c0\u03c9\u03c2 \u00ab\u03b2\u03c1\u03b5\u03c2 \u03c0\u03b1\u03c1\u03b1\u03b3\u03b3\u03b5\u03bb\u03af\u03b1, \u03ad\u03bb\u03b5\u03b3\u03be\u03b5 \u03c0\u03bb\u03b7\u03c1\u03c9\u03bc\u03ae, \u03b5\u03c0\u03b9\u03b2\u03b5\u03b2\u03b1\u03af\u03c9\u03c3\u03b5 \u03c0\u03bf\u03bb\u03b9\u03c4\u03b9\u03ba\u03ae, \u03ac\u03bb\u03bb\u03b1\u03be\u03b5 \u03ba\u03b1\u03c4\u03ac\u03c3\u03c4\u03b1\u03c3\u03b7, \u03b5\u03bd\u03b7\u03bc\u03ad\u03c1\u03c9\u03c3\u03b5 CRM \u03ba\u03b1\u03b9 \u03c3\u03c4\u03b5\u03af\u03bb\u03b5 \u03bc\u03ae\u03bd\u03c5\u03bc\u03b1\u00bb. \u0391\u03bd \u03ad\u03bd\u03b1 \u03bc\u03cc\u03bd\u03bf \u03b2\u03ae\u03bc\u03b1 \u03b5\u03ba\u03c4\u03b5\u03bb\u03b5\u03c3\u03c4\u03b5\u03af \u03c3\u03b5 \u03bb\u03ac\u03b8\u03bf\u03c2 record, \u03c4\u03bf \u03c4\u03b5\u03bb\u03b9\u03ba\u03cc \u03b1\u03c0\u03bf\u03c4\u03ad\u03bb\u03b5\u03c3\u03bc\u03b1 \u03bc\u03c0\u03bf\u03c1\u03b5\u03af \u03bd\u03b1 \u03c6\u03b1\u03af\u03bd\u03b5\u03c4\u03b1\u03b9 \u03bf\u03bb\u03bf\u03ba\u03bb\u03b7\u03c1\u03c9\u03bc\u03ad\u03bd\u03bf \u03b5\u03bd\u03ce \u03b5\u03af\u03bd\u03b1\u03b9 \u03b5\u03c0\u03b9\u03c7\u03b5\u03b9\u03c1\u03b7\u03c3\u03b9\u03b1\u03ba\u03ac \u03bb\u03b1\u03bd\u03b8\u03b1\u03c3\u03bc\u03ad\u03bd\u03bf.<\/p>\n\n<p>\u0397 \u03b1\u03c3\u03c6\u03b1\u03bb\u03ae\u03c2 \u03c3\u03c7\u03b5\u03b4\u03af\u03b1\u03c3\u03b7 \u03c3\u03c0\u03ac\u03b5\u03b9 \u03c4\u03b9\u03c2 \u03bc\u03b5\u03b3\u03ac\u03bb\u03b5\u03c2 \u03c1\u03bf\u03ad\u03c2 \u03c3\u03b5 \u03b5\u03c0\u03b1\u03bb\u03b7\u03b8\u03b5\u03cd\u03c3\u03b9\u03bc\u03b1 checkpoints. \u039a\u03ac\u03b8\u03b5 \u03c4\u03bc\u03ae\u03bc\u03b1 \u03c0\u03c1\u03ad\u03c0\u03b5\u03b9 \u03bd\u03b1 \u03ad\u03c7\u03b5\u03b9 precondition, \u03c3\u03c5\u03b3\u03ba\u03b5\u03ba\u03c1\u03b9\u03bc\u03ad\u03bd\u03bf \u03b1\u03c0\u03bf\u03c4\u03ad\u03bb\u03b5\u03c3\u03bc\u03b1, timeout, retry policy \u03ba\u03b1\u03b9 \u03c3\u03b1\u03c6\u03ae \u03b9\u03b4\u03b9\u03bf\u03ba\u03c4\u03ae\u03c4\u03b7. \u038c\u03c0\u03bf\u03c5 \u03b7 \u03b5\u03bd\u03ad\u03c1\u03b3\u03b5\u03b9\u03b1 \u03b5\u03af\u03bd\u03b1\u03b9 \u03bc\u03b7 \u03b1\u03bd\u03b1\u03c3\u03c4\u03c1\u03ad\u03c8\u03b9\u03bc\u03b7 \u03ae \u03b5\u03c0\u03b7\u03c1\u03b5\u03ac\u03b6\u03b5\u03b9 \u03c7\u03c1\u03ae\u03bc\u03b1\u03c4\u03b1, \u03b4\u03b9\u03ba\u03b1\u03b9\u03ce\u03bc\u03b1\u03c4\u03b1, \u03c0\u03b5\u03bb\u03ac\u03c4\u03b7 \u03ae \u03b4\u03b7\u03bc\u03bf\u03c3\u03af\u03b5\u03c5\u03c3\u03b7, \u03c7\u03c1\u03b5\u03b9\u03ac\u03b6\u03b5\u03c4\u03b1\u03b9 approval gate \u03c0\u03c1\u03b9\u03bd \u03c3\u03c5\u03bd\u03b5\u03c7\u03af\u03c3\u03b5\u03b9 \u03b7 \u03b1\u03bb\u03c5\u03c3\u03af\u03b4\u03b1.<\/p>\n\n<h2 id=\"epicheirimatiki-aksia\">\u03a4\u03b9 \u03c3\u03b7\u03bc\u03b1\u03af\u03bd\u03bf\u03c5\u03bd \u03c4\u03b1 \u03b5\u03c5\u03c1\u03ae\u03bc\u03b1\u03c4\u03b1 \u03b3\u03b9\u03b1 e-commerce \u03ba\u03b1\u03b9 \u03b5\u03c0\u03b9\u03c7\u03b5\u03b9\u03c1\u03b7\u03bc\u03b1\u03c4\u03b9\u03ba\u03ad\u03c2 \u03c1\u03bf\u03ad\u03c2<\/h2>\n\n<p>\u0393\u03b9\u03b1 \u03ad\u03bd\u03b1 e-shop, \u03bf\u03b9 \u03ba\u03b1\u03c4\u03ac\u03bb\u03bb\u03b7\u03bb\u03b5\u03c2 \u03c0\u03c1\u03ce\u03c4\u03b5\u03c2 \u03c7\u03c1\u03ae\u03c3\u03b5\u03b9\u03c2 \u03b4\u03b5\u03bd \u03b5\u03af\u03bd\u03b1\u03b9 \u03b7 \u03b1\u03c5\u03c4\u03cc\u03bd\u03bf\u03bc\u03b7 \u03b1\u03bb\u03bb\u03b1\u03b3\u03ae \u03c4\u03b9\u03bc\u03ce\u03bd \u03ae \u03b7 \u03b5\u03c0\u03b9\u03c3\u03c4\u03c1\u03bf\u03c6\u03ae \u03c7\u03c1\u03b7\u03bc\u03ac\u03c4\u03c9\u03bd \u03c7\u03c9\u03c1\u03af\u03c2 \u03ad\u03bb\u03b5\u03b3\u03c7\u03bf. \u0395\u03af\u03bd\u03b1\u03b9 \u03b5\u03c1\u03b3\u03b1\u03c3\u03af\u03b5\u03c2 \u03bc\u03b5 \u03c0\u03b5\u03c1\u03b9\u03bf\u03c1\u03b9\u03c3\u03bc\u03ad\u03bd\u03bf scope \u03ba\u03b1\u03b9 \u03b5\u03cd\u03ba\u03bf\u03bb\u03b7 \u03b5\u03c0\u03b1\u03bb\u03ae\u03b8\u03b5\u03c5\u03c3\u03b7: \u03c3\u03c5\u03b3\u03ba\u03ad\u03bd\u03c4\u03c1\u03c9\u03c3\u03b7 \u03c3\u03c4\u03bf\u03b9\u03c7\u03b5\u03af\u03c9\u03bd \u03b3\u03b9\u03b1 ticket, \u03c0\u03c1\u03cc\u03c4\u03b1\u03c3\u03b7 \u03b5\u03c0\u03cc\u03bc\u03b5\u03bd\u03b7\u03c2 \u03b5\u03bd\u03ad\u03c1\u03b3\u03b5\u03b9\u03b1\u03c2, \u03c0\u03c1\u03bf\u03b5\u03c4\u03bf\u03b9\u03bc\u03b1\u03c3\u03af\u03b1 \u03c0\u03c1\u03bf\u03c3\u03c7\u03b5\u03b4\u03af\u03bf\u03c5 \u03b1\u03c0\u03ac\u03bd\u03c4\u03b7\u03c3\u03b7\u03c2, \u03ad\u03bb\u03b5\u03b3\u03c7\u03bf\u03c2 \u03b1\u03bd \u03c3\u03c5\u03bc\u03c0\u03bb\u03b7\u03c1\u03ce\u03b8\u03b7\u03ba\u03b1\u03bd \u03cc\u03bb\u03b1 \u03c4\u03b1 \u03c5\u03c0\u03bf\u03c7\u03c1\u03b5\u03c9\u03c4\u03b9\u03ba\u03ac \u03c0\u03b5\u03b4\u03af\u03b1 \u03ae \u03b4\u03b7\u03bc\u03b9\u03bf\u03c5\u03c1\u03b3\u03af\u03b1 task \u03b3\u03b9\u03b1 \u03c4\u03bf\u03bd \u03c3\u03c9\u03c3\u03c4\u03cc \u03c5\u03c0\u03b5\u03cd\u03b8\u03c5\u03bd\u03bf. \u0397 \u03bf\u03bc\u03ac\u03b4\u03b1 \u03b4\u03b9\u03b1\u03c4\u03b7\u03c1\u03b5\u03af \u03c4\u03b7\u03bd \u03c4\u03b5\u03bb\u03b9\u03ba\u03ae \u03b1\u03c0\u03cc\u03c6\u03b1\u03c3\u03b7 \u03ba\u03b1\u03b9 \u03bf agent \u03bc\u03b5\u03b9\u03ce\u03bd\u03b5\u03b9 \u03c4\u03b7 \u03c7\u03b5\u03b9\u03c1\u03bf\u03ba\u03af\u03bd\u03b7\u03c4\u03b7 \u03b1\u03bd\u03b1\u03b6\u03ae\u03c4\u03b7\u03c3\u03b7.<\/p>\n\n<p>\u03a3\u03c4\u03bf customer support, \u03ad\u03bd\u03b1\u03c2 agent \u03bc\u03c0\u03bf\u03c1\u03b5\u03af \u03bd\u03b1 \u03c0\u03bb\u03bf\u03b7\u03b3\u03b5\u03af\u03c4\u03b1\u03b9 \u03c3\u03b5 WooCommerce, CRM \u03ba\u03b1\u03b9 help desk \u03b3\u03b9\u03b1 \u03bd\u03b1 \u03c3\u03c5\u03bd\u03b8\u03ad\u03c4\u03b5\u03b9 \u03c4\u03bf \u03b9\u03c3\u03c4\u03bf\u03c1\u03b9\u03ba\u03cc \u03b5\u03bd\u03cc\u03c2 \u03b1\u03b9\u03c4\u03ae\u03bc\u03b1\u03c4\u03bf\u03c2. \u0397 \u03b1\u03be\u03af\u03b1 \u03bc\u03b5\u03b3\u03b1\u03bb\u03ce\u03bd\u03b5\u03b9 \u03cc\u03c4\u03b1\u03bd \u03c4\u03bf \u03b1\u03c0\u03bf\u03c4\u03ad\u03bb\u03b5\u03c3\u03bc\u03b1 \u03c3\u03c5\u03bd\u03bf\u03b4\u03b5\u03cd\u03b5\u03c4\u03b1\u03b9 \u03b1\u03c0\u03cc \u03c3\u03b1\u03c6\u03ae \u03b4\u03b9\u03b1\u03b4\u03c1\u03bf\u03bc\u03ae: \u03c0\u03bf\u03b9\u03b1 \u03c0\u03b1\u03c1\u03b1\u03b3\u03b3\u03b5\u03bb\u03af\u03b1 \u03ac\u03bd\u03bf\u03b9\u03be\u03b5, \u03c0\u03bf\u03b9\u03b1 \u03c0\u03bf\u03bb\u03b9\u03c4\u03b9\u03ba\u03ae \u03b5\u03bb\u03ad\u03b3\u03c7\u03b8\u03b7\u03ba\u03b5 \u03ba\u03b1\u03b9 \u03c0\u03bf\u03b9\u03bf \u03c3\u03c4\u03bf\u03b9\u03c7\u03b5\u03af\u03bf \u03bb\u03b5\u03af\u03c0\u03b5\u03b9. \u0391\u03c5\u03c4\u03cc \u03c3\u03c5\u03bd\u03b4\u03ad\u03b5\u03c4\u03b1\u03b9 \u03ac\u03bc\u03b5\u03c3\u03b1 \u03bc\u03b5 \u03c4\u03b7\u03bd \u03b1\u03bd\u03ac\u03b3\u03ba\u03b7 \u03b3\u03b9\u03b1 <a href=\"https:\/\/twodots.gr\/oi-kalyteres-lyseis-help-desk-gia-epicheiriseis\/\">\u03bf\u03c1\u03b3\u03b1\u03bd\u03c9\u03bc\u03ad\u03bd\u03bf help desk \u03c3\u03b5 e-commerce \u03c1\u03bf\u03ad\u03c2<\/a>, \u03cc\u03c0\u03bf\u03c5 \u03c4\u03b1 \u03b4\u03b5\u03b4\u03bf\u03bc\u03ad\u03bd\u03b1 \u03c0\u03b5\u03bb\u03ac\u03c4\u03b7 \u03ba\u03b1\u03b9 \u03bf\u03b9 \u03b5\u03bd\u03ad\u03c1\u03b3\u03b5\u03b9\u03b5\u03c2 \u03c4\u03b7\u03c2 \u03bf\u03bc\u03ac\u03b4\u03b1\u03c2 \u03c0\u03c1\u03ad\u03c0\u03b5\u03b9 \u03bd\u03b1 \u03c0\u03b1\u03c1\u03b1\u03bc\u03ad\u03bd\u03bf\u03c5\u03bd \u03b5\u03bb\u03ad\u03b3\u03be\u03b9\u03bc\u03b1.<\/p>\n\n<p>\u03a3\u03c4\u03bf marketing, \u03bf\u03b9 web agents \u03bc\u03c0\u03bf\u03c1\u03bf\u03cd\u03bd \u03bd\u03b1 \u03c0\u03c1\u03bf\u03b5\u03c4\u03bf\u03b9\u03bc\u03ac\u03b6\u03bf\u03c5\u03bd reports, \u03bd\u03b1 \u03c3\u03c5\u03b3\u03ba\u03c1\u03af\u03bd\u03bf\u03c5\u03bd \u03c1\u03c5\u03b8\u03bc\u03af\u03c3\u03b5\u03b9\u03c2 \u03ae \u03bd\u03b1 \u03b5\u03bd\u03c4\u03bf\u03c0\u03af\u03b6\u03bf\u03c5\u03bd \u03b1\u03c3\u03c5\u03bd\u03ad\u03c0\u03b5\u03b9\u03b5\u03c2 \u03c0\u03c1\u03b9\u03bd \u03b1\u03c0\u03cc \u03b4\u03b7\u03bc\u03bf\u03c3\u03af\u03b5\u03c5\u03c3\u03b7. \u0394\u03b5\u03bd \u03c0\u03c1\u03ad\u03c0\u03b5\u03b9 \u03cc\u03bc\u03c9\u03c2 \u03bd\u03b1 \u03b1\u03c0\u03bf\u03ba\u03c4\u03bf\u03cd\u03bd \u03b1\u03bc\u03ad\u03c3\u03c9\u03c2 \u03b4\u03b9\u03ba\u03b1\u03af\u03c9\u03bc\u03b1 \u03bd\u03b1 \u03b1\u03bb\u03bb\u03ac\u03b6\u03bf\u03c5\u03bd budgets, audiences \u03ae tracking \u03c3\u03b5 \u03c0\u03bf\u03bb\u03bb\u03ac accounts. \u0397 \u03b1\u03c1\u03c7\u03ae \u03b5\u03af\u03bd\u03b1\u03b9 \u03af\u03b4\u03b9\u03b1 \u03bc\u03b5 \u03ba\u03ac\u03b8\u03b5 <a href=\"https:\/\/twodots.gr\/automation-robotics-ai-business\/\">\u03c0\u03c1\u03b1\u03ba\u03c4\u03b9\u03ba\u03ae \u03b5\u03c6\u03b1\u03c1\u03bc\u03bf\u03b3\u03ae automation \u03ba\u03b1\u03b9 AI<\/a>: \u03be\u03b5\u03ba\u03b9\u03bd\u03ac\u03bc\u03b5 \u03b1\u03c0\u03cc \u03bc\u03b5\u03c4\u03c1\u03ae\u03c3\u03b9\u03bc\u03bf bottleneck, \u03c0\u03b5\u03c1\u03b9\u03bf\u03c1\u03af\u03b6\u03bf\u03c5\u03bc\u03b5 \u03c4\u03bf action space \u03ba\u03b1\u03b9 \u03c0\u03b1\u03c1\u03b1\u03ba\u03bf\u03bb\u03bf\u03c5\u03b8\u03bf\u03cd\u03bc\u03b5 \u03c4\u03b7 \u03c1\u03bf\u03ae \u03bc\u03b5\u03c4\u03ac \u03c4\u03b7\u03bd \u03b5\u03bd\u03b5\u03c1\u03b3\u03bf\u03c0\u03bf\u03af\u03b7\u03c3\u03b7.<\/p>\n\n<p>\u03a4\u03bf MAG \u03c5\u03c0\u03b5\u03bd\u03b8\u03c5\u03bc\u03af\u03b6\u03b5\u03b9 \u03b5\u03c0\u03af\u03c3\u03b7\u03c2 \u03cc\u03c4\u03b9 \u03bf agent \u03ba\u03b1\u03b9 \u03b7 \u03c4\u03b5\u03ba\u03bc\u03b7\u03c1\u03af\u03c9\u03c3\u03b7 \u03bc\u03c0\u03bf\u03c1\u03bf\u03cd\u03bd \u03bd\u03b1 \u03c0\u03b1\u03c1\u03b1\u03c7\u03b8\u03bf\u03cd\u03bd \u03bc\u03b1\u03b6\u03af. \u03a3\u03b5 onboarding \u03b5\u03c1\u03b3\u03b1\u03b6\u03bf\u03bc\u03ad\u03bd\u03c9\u03bd, \u03b3\u03b9\u03b1 \u03c0\u03b1\u03c1\u03ac\u03b4\u03b5\u03b9\u03b3\u03bc\u03b1, \u03bc\u03b9\u03b1 \u03b5\u03c0\u03b1\u03bb\u03b7\u03b8\u03b5\u03c5\u03bc\u03ad\u03bd\u03b7 \u03b4\u03b9\u03b1\u03b4\u03c1\u03bf\u03bc\u03ae \u03bc\u03ad\u03c3\u03b1 \u03c3\u03b5 ERP \u03ae CRM \u03bc\u03c0\u03bf\u03c1\u03b5\u03af \u03bd\u03b1 \u03c4\u03c1\u03bf\u03c6\u03bf\u03b4\u03bf\u03c4\u03ae\u03c3\u03b5\u03b9 \u03bf\u03b4\u03b7\u03b3\u03af\u03b5\u03c2 \u03c0\u03bf\u03c5 \u03b1\u03bd\u03b1\u03bd\u03b5\u03ce\u03bd\u03bf\u03bd\u03c4\u03b1\u03b9 \u03c3\u03c5\u03c3\u03c4\u03b7\u03bc\u03b1\u03c4\u03b9\u03ba\u03ac. \u0391\u03c5\u03c4\u03cc \u03bc\u03b5\u03b9\u03ce\u03bd\u03b5\u03b9 \u03c4\u03b7 \u03b4\u03b9\u03b1\u03c6\u03bf\u03c1\u03ac \u03b1\u03bd\u03ac\u03bc\u03b5\u03c3\u03b1 \u03c3\u03c4\u03b7\u03bd \u03b1\u03c5\u03c4\u03bf\u03bc\u03b1\u03c4\u03bf\u03c0\u03bf\u03af\u03b7\u03c3\u03b7 \u03bc\u03b9\u03b1\u03c2 \u03b4\u03b9\u03b1\u03b4\u03b9\u03ba\u03b1\u03c3\u03af\u03b1\u03c2 \u03ba\u03b1\u03b9 \u03c3\u03c4\u03b7\u03bd <a href=\"https:\/\/twodots.gr\/pos-na-aftomatopoiisete-ti-diadikasia-ensomatosis-neon-ergazomenon\/\">\u03bf\u03c1\u03b3\u03ac\u03bd\u03c9\u03c3\u03b7 \u03c4\u03bf\u03c5 onboarding \u03b3\u03cd\u03c1\u03c9 \u03b1\u03c0\u03cc \u03c0\u03c1\u03b1\u03b3\u03bc\u03b1\u03c4\u03b9\u03ba\u03ac \u03c3\u03c5\u03c3\u03c4\u03ae\u03bc\u03b1\u03c4\u03b1<\/a>, \u03b1\u03c1\u03ba\u03b5\u03af \u03bf\u03b9 \u03bf\u03b4\u03b7\u03b3\u03af\u03b5\u03c2 \u03bd\u03b1 \u03c0\u03b5\u03c1\u03bd\u03bf\u03cd\u03bd \u03b1\u03bd\u03b8\u03c1\u03ce\u03c0\u03b9\u03bd\u03b7 \u03ad\u03b3\u03ba\u03c1\u03b9\u03c3\u03b7.<\/p>\n\n<h2 id=\"aksiologisi-dikou-agent\">\u03a0\u03ce\u03c2 \u03bd\u03b1 \u03b1\u03be\u03b9\u03bf\u03bb\u03bf\u03b3\u03ae\u03c3\u03b5\u03c4\u03b5 \u03c4\u03bf\u03bd \u03b4\u03b9\u03ba\u03cc \u03c3\u03b1\u03c2 AI agent<\/h2>\n\n<p>\u0388\u03bd\u03b1 \u03b3\u03b5\u03bd\u03b9\u03ba\u03cc benchmark \u03c0\u03c1\u03bf\u03c3\u03c6\u03ad\u03c1\u03b5\u03b9 \u03c3\u03b7\u03bc\u03b5\u03af\u03bf \u03b1\u03bd\u03b1\u03c6\u03bf\u03c1\u03ac\u03c2, \u03b1\u03bb\u03bb\u03ac \u03b4\u03b5\u03bd \u03b1\u03c0\u03b1\u03bd\u03c4\u03ac \u03b1\u03bd \u03bf agent \u03b5\u03af\u03bd\u03b1\u03b9 \u03b1\u03c3\u03c6\u03b1\u03bb\u03ae\u03c2 \u03c3\u03c4\u03bf \u03b4\u03b9\u03ba\u03cc \u03c3\u03b1\u03c2 WooCommerce, CRM, ERP \u03ae back office. \u0397 \u03c3\u03c9\u03c3\u03c4\u03ae \u03b1\u03be\u03b9\u03bf\u03bb\u03cc\u03b3\u03b7\u03c3\u03b7 \u03be\u03b5\u03ba\u03b9\u03bd\u03ac \u03b1\u03c0\u03cc \u03c0\u03c1\u03b1\u03b3\u03bc\u03b1\u03c4\u03b9\u03ba\u03ac tasks, \u03c0\u03c1\u03b1\u03b3\u03bc\u03b1\u03c4\u03b9\u03ba\u03ad\u03c2 \u03b5\u03be\u03b1\u03b9\u03c1\u03ad\u03c3\u03b5\u03b9\u03c2 \u03ba\u03b1\u03b9 \u03c4\u03bf \u03ba\u03cc\u03c3\u03c4\u03bf\u03c2 \u03b5\u03bd\u03cc\u03c2 \u03bb\u03ac\u03b8\u03bf\u03c5\u03c2. \u0391\u03ba\u03bf\u03bb\u03bf\u03c5\u03b8\u03b5\u03af \u03bc\u03b9\u03b1 \u03c0\u03c1\u03b1\u03ba\u03c4\u03b9\u03ba\u03ae \u03b4\u03b9\u03b1\u03b4\u03b9\u03ba\u03b1\u03c3\u03af\u03b1 \u03c0\u03bf\u03c5 \u03bc\u03b5\u03c4\u03b1\u03c6\u03ad\u03c1\u03b5\u03b9 \u03c4\u03b7 \u03bb\u03bf\u03b3\u03b9\u03ba\u03ae \u03c4\u03bf\u03c5 MAG \u03c3\u03b5 \u03b5\u03c4\u03b1\u03b9\u03c1\u03b9\u03ba\u03cc pilot.<\/p>\n\n\n<div class=\"td-step-list\"><p class=\"td-step-list-title\">\u0388\u03be\u03b9 \u03b2\u03ae\u03bc\u03b1\u03c4\u03b1 \u03b3\u03b9\u03b1 \u03b5\u03c0\u03b1\u03bb\u03b7\u03b8\u03b5\u03cd\u03c3\u03b9\u03bc\u03bf web-agent pilot<\/p><ol><li><span class=\"td-step-kicker\">Step 1<\/span><strong>\u039f\u03c1\u03af\u03c3\u03c4\u03b5 \u03ad\u03bd\u03b1 \u03c3\u03c4\u03b5\u03bd\u03cc \u03ba\u03b1\u03b9 \u03c7\u03c1\u03ae\u03c3\u03b9\u03bc\u03bf task<\/strong><p>\u0395\u03c0\u03b9\u03bb\u03ad\u03be\u03c4\u03b5 \u03bc\u03af\u03b1 \u03c1\u03bf\u03ae \u03bc\u03b5 \u03c3\u03b1\u03c6\u03ae \u03b1\u03c1\u03c7\u03ae \u03ba\u03b1\u03b9 \u03c4\u03ad\u03bb\u03bf\u03c2, \u03cc\u03c0\u03c9\u03c2 \u03c3\u03c5\u03bb\u03bb\u03bf\u03b3\u03ae \u03c3\u03c4\u03bf\u03b9\u03c7\u03b5\u03af\u03c9\u03bd \u03c0\u03b1\u03c1\u03b1\u03b3\u03b3\u03b5\u03bb\u03af\u03b1\u03c2 \u03b3\u03b9\u03b1 ticket, \u03ba\u03b1\u03b9 \u03bf\u03c1\u03af\u03c3\u03c4\u03b5 \u03c0\u03bf\u03b9\u03b1 \u03b5\u03bd\u03ad\u03c1\u03b3\u03b5\u03b9\u03b1 \u03bc\u03ad\u03bd\u03b5\u03b9 \u03b1\u03c0\u03bf\u03ba\u03bb\u03b5\u03b9\u03c3\u03c4\u03b9\u03ba\u03ac \u03c3\u03c4\u03bf\u03bd \u03ac\u03bd\u03b8\u03c1\u03c9\u03c0\u03bf.<\/p><\/li><li><span class=\"td-step-kicker\">Step 2<\/span><strong>\u039a\u03b1\u03c4\u03b1\u03b3\u03c1\u03ac\u03c8\u03c4\u03b5 \u03c4\u03b7\u03bd \u03b1\u03c1\u03c7\u03b9\u03ba\u03ae \u03ba\u03b1\u03c4\u03ac\u03c3\u03c4\u03b1\u03c3\u03b7<\/strong><p>\u039f\u03c1\u03af\u03c3\u03c4\u03b5 \u03bb\u03bf\u03b3\u03b1\u03c1\u03b9\u03b1\u03c3\u03bc\u03cc, \u03b4\u03b9\u03ba\u03b1\u03b9\u03ce\u03bc\u03b1\u03c4\u03b1, viewport, \u03b4\u03b5\u03b4\u03bf\u03bc\u03ad\u03bd\u03b1, \u03b1\u03bd\u03bf\u03b9\u03c7\u03c4\u03ad\u03c2 \u03ba\u03b1\u03c1\u03c4\u03ad\u03bb\u03b5\u03c2 \u03ba\u03b1\u03b9 \u03c0\u03c1\u03bf\u03cb\u03c0\u03bf\u03b8\u03ad\u03c3\u03b5\u03b9\u03c2, \u03ce\u03c3\u03c4\u03b5 \u03ba\u03ac\u03b8\u03b5 \u03b5\u03c0\u03b1\u03bd\u03ac\u03bb\u03b7\u03c8\u03b7 \u03bd\u03b1 \u03be\u03b5\u03ba\u03b9\u03bd\u03ac \u03b1\u03c0\u03cc \u03c3\u03c5\u03b3\u03ba\u03c1\u03af\u03c3\u03b9\u03bc\u03bf \u03c0\u03b5\u03c1\u03b9\u03b2\u03ac\u03bb\u03bb\u03bf\u03bd.<\/p><\/li><li><span class=\"td-step-kicker\">Step 3<\/span><strong>\u03a7\u03c4\u03af\u03c3\u03c4\u03b5 \u03b1\u03bd\u03c4\u03b9\u03ba\u03b5\u03b9\u03bc\u03b5\u03bd\u03b9\u03ba\u03cc checker<\/strong><p>\u0395\u03bb\u03ad\u03b3\u03be\u03c4\u03b5 \u03c4\u03bf \u03c0\u03c1\u03b1\u03b3\u03bc\u03b1\u03c4\u03b9\u03ba\u03cc \u03c4\u03b5\u03bb\u03b9\u03ba\u03cc state \u03bc\u03ad\u03c3\u03c9 API \u03ae \u03b2\u03ac\u03c3\u03b7\u03c2 \u03cc\u03c0\u03bf\u03c5 \u03b3\u03af\u03bd\u03b5\u03c4\u03b1\u03b9, \u03cc\u03c7\u03b9 \u03bc\u03cc\u03bd\u03bf \u03c4\u03bf \u03bc\u03ae\u03bd\u03c5\u03bc\u03b1 \u00ab\u03bf\u03bb\u03bf\u03ba\u03bb\u03b7\u03c1\u03ce\u03b8\u03b7\u03ba\u03b5\u00bb \u03c0\u03bf\u03c5 \u03b5\u03c0\u03b9\u03c3\u03c4\u03c1\u03ad\u03c6\u03b5\u03b9 \u03bf agent.<\/p><\/li><li><span class=\"td-step-kicker\">Step 4<\/span><strong>\u0392\u03b1\u03b8\u03bc\u03bf\u03bb\u03bf\u03b3\u03ae\u03c3\u03c4\u03b5 \u03b5\u03bd\u03ad\u03c1\u03b3\u03b5\u03b9\u03b1 \u03ba\u03b1\u03b9 \u03b5\u03be\u03ae\u03b3\u03b7\u03c3\u03b7 \u03c7\u03c9\u03c1\u03b9\u03c3\u03c4\u03ac<\/strong><p>\u039c\u03b5\u03c4\u03c1\u03ae\u03c3\u03c4\u03b5 task success, \u03bb\u03ac\u03b8\u03bf\u03c2 clicks, retries, \u03c7\u03c1\u03cc\u03bd\u03bf, \u03ba\u03cc\u03c3\u03c4\u03bf\u03c2 \u03ba\u03b1\u03b9 \u03b1\u03bd \u03bf\u03b9 \u03bf\u03b4\u03b7\u03b3\u03af\u03b5\u03c2 \u03c0\u03b5\u03c1\u03b9\u03b3\u03c1\u03ac\u03c6\u03bf\u03c5\u03bd \u03c3\u03c9\u03c3\u03c4\u03ac \u03c4\u03b7\u03bd \u03bf\u03c1\u03b1\u03c4\u03ae \u03ba\u03b1\u03c4\u03ac\u03c3\u03c4\u03b1\u03c3\u03b7 \u03c7\u03c9\u03c1\u03af\u03c2 \u03c4\u03b5\u03c7\u03bd\u03b9\u03ba\u03ae \u03b4\u03b9\u03b1\u03c1\u03c1\u03bf\u03ae.<\/p><\/li><li><span class=\"td-step-kicker\">Step 5<\/span><strong>\u0391\u03bd\u03b1\u03bb\u03cd\u03c3\u03c4\u03b5 \u03bc\u03ae\u03ba\u03bf\u03c2 \u03ba\u03b1\u03b9 severity<\/strong><p>\u03a7\u03c9\u03c1\u03af\u03c3\u03c4\u03b5 \u03c4\u03b1 tests \u03b1\u03bd\u03ac \u03b1\u03c1\u03b9\u03b8\u03bc\u03cc \u03b2\u03b7\u03bc\u03ac\u03c4\u03c9\u03bd \u03ba\u03b1\u03b9 \u03b5\u03c0\u03af\u03c0\u03c4\u03c9\u03c3\u03b7. \u0388\u03bd\u03b1 \u03bb\u03ac\u03b8\u03bf\u03c2 \u03c3\u03b5 \u03c0\u03c1\u03bf\u03c3\u03c7\u03ad\u03b4\u03b9\u03bf \u03b1\u03bd\u03b1\u03c6\u03bf\u03c1\u03ac\u03c2 \u03b4\u03b5\u03bd \u03ad\u03c7\u03b5\u03b9 \u03c4\u03bf \u03af\u03b4\u03b9\u03bf \u03c1\u03af\u03c3\u03ba\u03bf \u03bc\u03b5 \u03b1\u03bb\u03bb\u03b1\u03b3\u03ae \u03c0\u03b1\u03c1\u03b1\u03b3\u03b3\u03b5\u03bb\u03af\u03b1\u03c2 \u03ae \u03b1\u03c0\u03bf\u03c3\u03c4\u03bf\u03bb\u03ae \u03bc\u03b7\u03bd\u03cd\u03bc\u03b1\u03c4\u03bf\u03c2 \u03c3\u03b5 \u03c0\u03b5\u03bb\u03ac\u03c4\u03b7.<\/p><\/li><li><span class=\"td-step-kicker\">Step 6<\/span><strong>\u03a4\u03c1\u03ad\u03be\u03c4\u03b5 shadow mode \u03ba\u03b1\u03b9 regression<\/strong><p>\u0391\u03c6\u03ae\u03c3\u03c4\u03b5 \u03c4\u03bf\u03bd agent \u03bd\u03b1 \u03c0\u03c1\u03bf\u03c4\u03b5\u03af\u03bd\u03b5\u03b9 \u03c7\u03c9\u03c1\u03af\u03c2 \u03bd\u03b1 \u03b5\u03ba\u03c4\u03b5\u03bb\u03b5\u03af, \u03c3\u03c5\u03b3\u03ba\u03c1\u03af\u03bd\u03b5\u03c4\u03b5 \u03bc\u03b5 \u03b1\u03bd\u03b8\u03c1\u03ce\u03c0\u03b9\u03bd\u03b7 \u03b1\u03c0\u03cc\u03c6\u03b1\u03c3\u03b7 \u03ba\u03b1\u03b9 \u03b5\u03c0\u03b1\u03bd\u03b1\u03bb\u03ac\u03b2\u03b5\u03c4\u03b5 \u03c4\u03bf test set \u03c3\u03b5 \u03ba\u03ac\u03b8\u03b5 \u03b1\u03bb\u03bb\u03b1\u03b3\u03ae \u03bc\u03bf\u03bd\u03c4\u03ad\u03bb\u03bf\u03c5, prompt \u03ae website.<\/p><\/li><\/ol><\/div>\n\n\n<p>\u0397 \u03b1\u03be\u03b9\u03bf\u03bb\u03cc\u03b3\u03b7\u03c3\u03b7 \u03c7\u03c1\u03b5\u03b9\u03ac\u03b6\u03b5\u03c4\u03b1\u03b9 \u03ba\u03b1\u03b9 \u03ba\u03c1\u03b9\u03c4\u03ae\u03c1\u03b9\u03bf \u03b1\u03c0\u03bf\u03c7\u03ae\u03c2. \u0388\u03bd\u03b1\u03c2 \u03b1\u03be\u03b9\u03cc\u03c0\u03b9\u03c3\u03c4\u03bf\u03c2 agent \u03b4\u03b5\u03bd \u03c0\u03c1\u03ad\u03c0\u03b5\u03b9 \u03bd\u03b1 \u03c0\u03c1\u03bf\u03c3\u03c0\u03b1\u03b8\u03b5\u03af \u03c0\u03ac\u03bd\u03c4\u03b1. \u038c\u03c4\u03b1\u03bd \u03bb\u03b5\u03af\u03c0\u03b5\u03b9 \u03c3\u03c4\u03bf\u03b9\u03c7\u03b5\u03af\u03bf, \u03b1\u03bb\u03bb\u03ac\u03b6\u03b5\u03b9 \u03b7 \u03c3\u03b5\u03bb\u03af\u03b4\u03b1, \u03b1\u03c0\u03bf\u03c4\u03c5\u03b3\u03c7\u03ac\u03bd\u03b5\u03b9 \u03bf checker \u03ae \u03b7 \u03b5\u03bd\u03ad\u03c1\u03b3\u03b5\u03b9\u03b1 \u03be\u03b5\u03c0\u03b5\u03c1\u03bd\u03ac \u03c4\u03b1 \u03b4\u03b9\u03ba\u03b1\u03b9\u03ce\u03bc\u03b1\u03c4\u03ac \u03c4\u03bf\u03c5, \u03b7 \u03c3\u03c9\u03c3\u03c4\u03ae \u03c3\u03c5\u03bc\u03c0\u03b5\u03c1\u03b9\u03c6\u03bf\u03c1\u03ac \u03b5\u03af\u03bd\u03b1\u03b9 \u03bd\u03b1 \u03c3\u03c4\u03b1\u03bc\u03b1\u03c4\u03ac \u03ba\u03b1\u03b9 \u03bd\u03b1 \u03b6\u03b7\u03c4\u03ac \u03ac\u03bd\u03b8\u03c1\u03c9\u03c0\u03bf. \u0397 <a href=\"https:\/\/twodots.gr\/ai-agents-confidence-context-data-human-oversight\/\">\u03b1\u03bd\u03b8\u03c1\u03ce\u03c0\u03b9\u03bd\u03b7 \u03b5\u03c0\u03bf\u03c0\u03c4\u03b5\u03af\u03b1 \u03c3\u03c4\u03bf\u03c5\u03c2 AI agents<\/a> \u03b4\u03b5\u03bd \u03b5\u03af\u03bd\u03b1\u03b9 \u03b5\u03c6\u03b5\u03b4\u03c1\u03b9\u03ba\u03ae \u03bb\u03b5\u03b9\u03c4\u03bf\u03c5\u03c1\u03b3\u03af\u03b1\u00b7 \u03b5\u03af\u03bd\u03b1\u03b9 \u03bc\u03ad\u03c1\u03bf\u03c2 \u03c4\u03b7\u03c2 \u03b1\u03c1\u03c7\u03b9\u03c4\u03b5\u03ba\u03c4\u03bf\u03bd\u03b9\u03ba\u03ae\u03c2 \u03c4\u03bf\u03c5 workflow.<\/p>\n\n<h2 id=\"production-guardrails\">\u03a4\u03b1 guardrails \u03c0\u03c1\u03b9\u03bd \u03b1\u03c0\u03cc \u03c4\u03b7\u03bd \u03c0\u03b1\u03c1\u03b1\u03b3\u03c9\u03b3\u03b9\u03ba\u03ae \u03c7\u03c1\u03ae\u03c3\u03b7<\/h2>\n\n<p>\u03a4\u03bf \u03c0\u03c1\u03ce\u03c4\u03bf guardrail \u03b5\u03af\u03bd\u03b1\u03b9 \u03c4\u03bf least privilege. \u039f agent \u03c0\u03c1\u03ad\u03c0\u03b5\u03b9 \u03bd\u03b1 \u03b2\u03bb\u03ad\u03c0\u03b5\u03b9 \u03ba\u03b1\u03b9 \u03bd\u03b1 \u03b1\u03bb\u03bb\u03ac\u03b6\u03b5\u03b9 \u03bc\u03cc\u03bd\u03bf \u03cc\u03c3\u03b1 \u03b1\u03c0\u03b1\u03b9\u03c4\u03b5\u03af \u03c4\u03bf \u03c3\u03c5\u03b3\u03ba\u03b5\u03ba\u03c1\u03b9\u03bc\u03ad\u03bd\u03bf task. \u039e\u03b5\u03c7\u03c9\u03c1\u03b9\u03c3\u03c4\u03cc\u03c2 \u03bb\u03bf\u03b3\u03b1\u03c1\u03b9\u03b1\u03c3\u03bc\u03cc\u03c2, \u03c0\u03b5\u03c1\u03b9\u03bf\u03c1\u03b9\u03c3\u03bc\u03ad\u03bd\u03b1 roles \u03ba\u03b1\u03b9 \u03c0\u03b5\u03c1\u03b9\u03b2\u03ac\u03bb\u03bb\u03bf\u03bd staging \u03bc\u03b5\u03b9\u03ce\u03bd\u03bf\u03c5\u03bd \u03c4\u03b7\u03bd \u03ad\u03ba\u03c4\u03b1\u03c3\u03b7 \u03b5\u03bd\u03cc\u03c2 \u03bb\u03ac\u03b8\u03bf\u03c5\u03c2. \u03a4\u03bf \u03b4\u03b5\u03cd\u03c4\u03b5\u03c1\u03bf \u03b5\u03af\u03bd\u03b1\u03b9 \u03b7 \u03b1\u03bd\u03b1\u03c3\u03c4\u03c1\u03b5\u03c8\u03b9\u03bc\u03cc\u03c4\u03b7\u03c4\u03b1: drafts, \u03c0\u03c1\u03bf\u03c4\u03ac\u03c3\u03b5\u03b9\u03c2 \u03ba\u03b1\u03b9 \u03c0\u03c1\u03bf\u03c3\u03c9\u03c1\u03b9\u03bd\u03ad\u03c2 \u03ba\u03b1\u03c4\u03b1\u03c3\u03c4\u03ac\u03c3\u03b5\u03b9\u03c2 \u03c0\u03c1\u03bf\u03b7\u03b3\u03bf\u03cd\u03bd\u03c4\u03b1\u03b9 \u03c4\u03c9\u03bd \u03bf\u03c1\u03b9\u03c3\u03c4\u03b9\u03ba\u03ce\u03bd \u03b5\u03bd\u03b5\u03c1\u03b3\u03b5\u03b9\u03ce\u03bd.<\/p>\n\n<p>\u03a4\u03bf \u03c4\u03c1\u03af\u03c4\u03bf \u03b5\u03af\u03bd\u03b1\u03b9 \u03b7 \u03b1\u03bd\u03b5\u03be\u03ac\u03c1\u03c4\u03b7\u03c4\u03b7 \u03b5\u03c0\u03b1\u03bb\u03ae\u03b8\u03b5\u03c5\u03c3\u03b7. \u03a4\u03bf \u03af\u03b4\u03b9\u03bf \u03bc\u03bf\u03bd\u03c4\u03ad\u03bb\u03bf \u03c0\u03bf\u03c5 \u03ad\u03ba\u03b1\u03bd\u03b5 \u03c4\u03b7\u03bd \u03b5\u03bd\u03ad\u03c1\u03b3\u03b5\u03b9\u03b1 \u03b4\u03b5\u03bd \u03c0\u03c1\u03ad\u03c0\u03b5\u03b9 \u03bd\u03b1 \u03b5\u03af\u03bd\u03b1\u03b9 \u03bf \u03bc\u03bf\u03bd\u03b1\u03b4\u03b9\u03ba\u03cc\u03c2 \u03ba\u03c1\u03b9\u03c4\u03ae\u03c2 \u03c4\u03b7\u03c2 \u03b5\u03c0\u03b9\u03c4\u03c5\u03c7\u03af\u03b1\u03c2 \u03c4\u03bf\u03c5. \u03a7\u03c1\u03b7\u03c3\u03b9\u03bc\u03bf\u03c0\u03bf\u03b9\u03ae\u03c3\u03c4\u03b5 API readback, \u03ba\u03b1\u03bd\u03cc\u03bd\u03b5\u03c2, database state \u03ae \u03b4\u03b5\u03cd\u03c4\u03b5\u03c1\u03bf \u03b5\u03bb\u03b5\u03b3\u03ba\u03c4\u03b9\u03ba\u03cc \u03bc\u03bf\u03bd\u03bf\u03c0\u03ac\u03c4\u03b9. \u03a4\u03bf \u03c4\u03ad\u03c4\u03b1\u03c1\u03c4\u03bf \u03b5\u03af\u03bd\u03b1\u03b9 \u03b7 \u03c0\u03b1\u03c1\u03b1\u03c4\u03b7\u03c1\u03b7\u03c3\u03b9\u03bc\u03cc\u03c4\u03b7\u03c4\u03b1: screenshots, timestamps, action log, input version, model version, \u03ba\u03cc\u03c3\u03c4\u03bf\u03c2 \u03ba\u03b1\u03b9 reason for escalation \u03c0\u03c1\u03ad\u03c0\u03b5\u03b9 \u03bd\u03b1 \u03bc\u03c0\u03bf\u03c1\u03bf\u03cd\u03bd \u03bd\u03b1 \u03b1\u03bd\u03b1\u03c3\u03c5\u03bd\u03c4\u03b5\u03b8\u03bf\u03cd\u03bd.<\/p>\n\n<p>\u03a4\u03ad\u03bb\u03bf\u03c2 \u03c7\u03c1\u03b5\u03b9\u03ac\u03b6\u03b5\u03c4\u03b1\u03b9 risk owner. \u039a\u03ac\u03c0\u03bf\u03b9\u03bf\u03c2 \u03c0\u03c1\u03ad\u03c0\u03b5\u03b9 \u03bd\u03b1 \u03b1\u03c0\u03bf\u03c6\u03b1\u03c3\u03af\u03b6\u03b5\u03b9 \u03c0\u03bf\u03b9\u03b1 failure rate \u03b5\u03af\u03bd\u03b1\u03b9 \u03b1\u03c0\u03bf\u03b4\u03b5\u03ba\u03c4\u03ae, \u03c0\u03bf\u03b9\u03b1 tasks \u03b1\u03c0\u03b1\u03b3\u03bf\u03c1\u03b5\u03cd\u03bf\u03bd\u03c4\u03b1\u03b9, \u03c0\u03cc\u03c4\u03b5 \u03b3\u03af\u03bd\u03b5\u03c4\u03b1\u03b9 rollback \u03ba\u03b1\u03b9 \u03c0\u03ce\u03c2 \u03b5\u03bd\u03b7\u03bc\u03b5\u03c1\u03ce\u03bd\u03b5\u03c4\u03b1\u03b9 \u03bf \u03c0\u03b5\u03bb\u03ac\u03c4\u03b7\u03c2 \u03ae \u03b7 \u03bf\u03bc\u03ac\u03b4\u03b1. \u03a4\u03bf NIST AI RMF \u03bf\u03c1\u03b3\u03b1\u03bd\u03ce\u03bd\u03b5\u03b9 \u03b1\u03c5\u03c4\u03ae \u03c4\u03b7 \u03bb\u03bf\u03b3\u03b9\u03ba\u03ae \u03c3\u03c4\u03b9\u03c2 \u03bb\u03b5\u03b9\u03c4\u03bf\u03c5\u03c1\u03b3\u03af\u03b5\u03c2 Govern, Map, Measure \u03ba\u03b1\u03b9 Manage. \u0393\u03b9\u03b1 \u03ad\u03bd\u03b1\u03bd web agent, \u03b1\u03c5\u03c4\u03ad\u03c2 \u03bc\u03b5\u03c4\u03b1\u03c6\u03c1\u03ac\u03b6\u03bf\u03bd\u03c4\u03b1\u03b9 \u03c3\u03b5 \u03b9\u03b4\u03b9\u03bf\u03ba\u03c4\u03b7\u03c3\u03af\u03b1, \u03c7\u03b1\u03c1\u03c4\u03bf\u03b3\u03c1\u03ac\u03c6\u03b7\u03c3\u03b7 use case, \u03bc\u03b5\u03c4\u03c1\u03ae\u03c3\u03b5\u03b9\u03c2 \u03c3\u03b5 \u03c0\u03c1\u03b1\u03b3\u03bc\u03b1\u03c4\u03b9\u03ba\u03ad\u03c2 \u03c1\u03bf\u03ad\u03c2 \u03ba\u03b1\u03b9 \u03b5\u03bd\u03b5\u03c1\u03b3\u03ac controls \u03ba\u03b1\u03c4\u03ac \u03c4\u03b7 \u03bb\u03b5\u03b9\u03c4\u03bf\u03c5\u03c1\u03b3\u03af\u03b1.<\/p>\n\n<h2 id=\"mag-kritirio-agoras\">\u03a4\u03bf MAG \u03b1\u03bb\u03bb\u03ac\u03b6\u03b5\u03b9 \u03c4\u03bf \u03b2\u03b1\u03c3\u03b9\u03ba\u03cc \u03ba\u03c1\u03b9\u03c4\u03ae\u03c1\u03b9\u03bf \u03b1\u03b3\u03bf\u03c1\u03ac\u03c2 \u03b5\u03bd\u03cc\u03c2 AI agent<\/h2>\n\n<p>\u0397 \u03c3\u03c9\u03c3\u03c4\u03ae \u03b5\u03c1\u03ce\u03c4\u03b7\u03c3\u03b7 \u03b4\u03b5\u03bd \u03b5\u03af\u03bd\u03b1\u03b9 \u00ab\u03bc\u03c0\u03bf\u03c1\u03b5\u03af \u03bd\u03b1 \u03c7\u03c1\u03b7\u03c3\u03b9\u03bc\u03bf\u03c0\u03bf\u03b9\u03ae\u03c3\u03b5\u03b9 browser;\u00bb. \u0395\u03af\u03bd\u03b1\u03b9 \u00ab\u03c3\u03b5 \u03c0\u03bf\u03b9\u03b1 \u03b4\u03b9\u03ba\u03ac \u03bc\u03b1\u03c2 tasks \u03bf\u03bb\u03bf\u03ba\u03bb\u03b7\u03c1\u03ce\u03bd\u03b5\u03b9 \u03b5\u03c0\u03b1\u03bb\u03b7\u03b8\u03b5\u03cd\u03c3\u03b9\u03bc\u03b1 \u03c4\u03bf\u03bd \u03c3\u03c4\u03cc\u03c7\u03bf, \u03bc\u03b5 \u03c0\u03cc\u03c3\u03b1 \u03b2\u03ae\u03bc\u03b1\u03c4\u03b1, \u03c0\u03bf\u03b9\u03b1 \u03bb\u03ac\u03b8\u03b7, \u03c4\u03b9 \u03b5\u03be\u03ae\u03b3\u03b7\u03c3\u03b7 \u03ba\u03b1\u03b9 \u03c0\u03bf\u03b9\u03bf fallback;\u00bb. \u03a4\u03bf MAG \u03b2\u03bf\u03b7\u03b8\u03ac \u03bd\u03b1 \u03bc\u03b5\u03c4\u03b1\u03ba\u03b9\u03bd\u03b7\u03b8\u03b5\u03af \u03b7 \u03c3\u03c5\u03b6\u03ae\u03c4\u03b7\u03c3\u03b7 \u03b1\u03c0\u03cc \u03c4\u03b7\u03bd \u03b5\u03bd\u03c4\u03cd\u03c0\u03c9\u03c3\u03b7 \u03c4\u03bf\u03c5 demo \u03c3\u03b5 \u03bc\u03b5\u03c4\u03c1\u03ae\u03c3\u03b5\u03b9\u03c2 \u03c0\u03bf\u03c5 \u03c3\u03c5\u03bd\u03b4\u03ad\u03bf\u03bd\u03c4\u03b1\u03b9 \u03bc\u03b5 \u03c0\u03c1\u03b1\u03b3\u03bc\u03b1\u03c4\u03b9\u03ba\u03ae \u03c7\u03c1\u03ae\u03c3\u03b7.<\/p>\n\n<p>\u0388\u03bd\u03b1\u03c2 vendor \u03c0\u03c1\u03ad\u03c0\u03b5\u03b9 \u03bd\u03b1 \u03bc\u03c0\u03bf\u03c1\u03b5\u03af \u03bd\u03b1 \u03b4\u03b5\u03af\u03be\u03b5\u03b9 \u03b1\u03c0\u03bf\u03c4\u03b5\u03bb\u03ad\u03c3\u03bc\u03b1\u03c4\u03b1 \u03b1\u03bd\u03ac task length, grounding method, website, model version \u03ba\u03b1\u03b9 \u03b5\u03c0\u03af\u03c0\u03b5\u03b4\u03bf \u03c1\u03af\u03c3\u03ba\u03bf\u03c5. \u03a0\u03c1\u03ad\u03c0\u03b5\u03b9 \u03b5\u03c0\u03af\u03c3\u03b7\u03c2 \u03bd\u03b1 \u03b5\u03be\u03b7\u03b3\u03b5\u03af \u03c0\u03ce\u03c2 \u03b5\u03bb\u03ad\u03b3\u03c7\u03b5\u03c4\u03b1\u03b9 \u03c4\u03bf finish, \u03c0\u03ce\u03c2 \u03b1\u03c0\u03bf\u03c4\u03c1\u03ad\u03c0\u03b5\u03c4\u03b1\u03b9 \u03b7 \u03b5\u03bd\u03ad\u03c1\u03b3\u03b5\u03b9\u03b1 \u03c3\u03b5 \u03bb\u03ac\u03b8\u03bf\u03c2 record \u03ba\u03b1\u03b9 \u03c0\u03ce\u03c2 \u03b1\u03bd\u03b1\u03bd\u03b5\u03ce\u03bd\u03bf\u03bd\u03c4\u03b1\u03b9 \u03bf\u03b9 \u03bf\u03b4\u03b7\u03b3\u03af\u03b5\u03c2 \u03cc\u03c4\u03b1\u03bd \u03b1\u03bb\u03bb\u03ac\u03be\u03b5\u03b9 \u03c4\u03bf UI. \u0391\u03bd \u03bf\u03b9 \u03b1\u03c0\u03b1\u03bd\u03c4\u03ae\u03c3\u03b5\u03b9\u03c2 \u03c0\u03b5\u03c1\u03b9\u03bf\u03c1\u03af\u03b6\u03bf\u03bd\u03c4\u03b1\u03b9 \u03c3\u03b5 \u03c3\u03c5\u03bd\u03bf\u03bb\u03b9\u03ba\u03cc success rate \u03ae \u03c3\u03b5 \u03b2\u03af\u03bd\u03c4\u03b5\u03bf \u03b5\u03c0\u03b9\u03bb\u03b5\u03b3\u03bc\u03ad\u03bd\u03c9\u03bd \u03b4\u03b9\u03b1\u03b4\u03c1\u03bf\u03bc\u03ce\u03bd, \u03c4\u03bf \u03bb\u03b5\u03b9\u03c4\u03bf\u03c5\u03c1\u03b3\u03b9\u03ba\u03cc \u03c1\u03af\u03c3\u03ba\u03bf \u03c0\u03b1\u03c1\u03b1\u03bc\u03ad\u03bd\u03b5\u03b9 \u03ac\u03b3\u03bd\u03c9\u03c3\u03c4\u03bf.<\/p>\n\n<p>\u0393\u03b9\u03b1 \u03c4\u03b9\u03c2 \u03b5\u03c0\u03b9\u03c7\u03b5\u03b9\u03c1\u03ae\u03c3\u03b5\u03b9\u03c2, \u03b7 \u03c0\u03b9\u03bf \u03c1\u03b5\u03b1\u03bb\u03b9\u03c3\u03c4\u03b9\u03ba\u03ae \u03c0\u03bf\u03c1\u03b5\u03af\u03b1 \u03b5\u03af\u03bd\u03b1\u03b9 \u03c3\u03c4\u03b1\u03b4\u03b9\u03b1\u03ba\u03ae: \u03bc\u03b9\u03ba\u03c1\u03cc task, \u03b1\u03bd\u03c4\u03b9\u03ba\u03b5\u03b9\u03bc\u03b5\u03bd\u03b9\u03ba\u03cc\u03c2 checker, shadow mode, approval gates \u03ba\u03b1\u03b9 \u03b5\u03c0\u03ad\u03ba\u03c4\u03b1\u03c3\u03b7 \u03bc\u03cc\u03bd\u03bf \u03cc\u03c4\u03b1\u03bd \u03c4\u03b1 \u03b4\u03b5\u03b4\u03bf\u03bc\u03ad\u03bd\u03b1 \u03b4\u03b5\u03af\u03c7\u03bd\u03bf\u03c5\u03bd \u03c3\u03c4\u03b1\u03b8\u03b5\u03c1\u03cc\u03c4\u03b7\u03c4\u03b1. \u039f\u03b9 web agents \u03bc\u03c0\u03bf\u03c1\u03bf\u03cd\u03bd \u03bd\u03b1 \u03bc\u03b5\u03b9\u03ce\u03c3\u03bf\u03c5\u03bd \u03c7\u03b5\u03b9\u03c1\u03bf\u03ba\u03af\u03bd\u03b7\u03c4\u03b1 \u03b2\u03ae\u03bc\u03b1\u03c4\u03b1 \u03ba\u03b1\u03b9 \u03bd\u03b1 \u03bc\u03b5\u03c4\u03b1\u03c4\u03c1\u03ad\u03c8\u03bf\u03c5\u03bd \u03b5\u03c0\u03b9\u03c4\u03c5\u03c7\u03b7\u03bc\u03ad\u03bd\u03b5\u03c2 \u03b4\u03b9\u03b1\u03b4\u03c1\u03bf\u03bc\u03ad\u03c2 \u03c3\u03b5 \u03c7\u03c1\u03ae\u03c3\u03b9\u03bc\u03b7 \u03ba\u03b1\u03b8\u03bf\u03b4\u03ae\u03b3\u03b7\u03c3\u03b7. \u03a4\u03bf MAG \u03cc\u03bc\u03c9\u03c2 \u03b4\u03b5\u03af\u03c7\u03bd\u03b5\u03b9 \u03ba\u03b1\u03b8\u03b1\u03c1\u03ac \u03cc\u03c4\u03b9 \u03b7 \u03b1\u03be\u03af\u03b1 \u03c4\u03bf\u03c5\u03c2 \u03b1\u03c1\u03c7\u03af\u03b6\u03b5\u03b9 \u03b5\u03ba\u03b5\u03af \u03cc\u03c0\u03bf\u03c5 \u03c4\u03b5\u03bb\u03b5\u03b9\u03ce\u03bd\u03b5\u03b9 \u03c4\u03bf demo \u03ba\u03b1\u03b9 \u03be\u03b5\u03ba\u03b9\u03bd\u03ac \u03b7 \u03c3\u03c5\u03c3\u03c4\u03b7\u03bc\u03b1\u03c4\u03b9\u03ba\u03ae \u03b5\u03c0\u03b1\u03bb\u03ae\u03b8\u03b5\u03c5\u03c3\u03b7.<\/p>\n\n\n<section class=\"td-service-cta\"><div class=\"td-service-cta-content\"><p class=\"td-service-cta-eyebrow\">\u0391\u03c0\u03cc \u03c4\u03bf web-agent demo \u03c3\u03b5 \u03b5\u03bb\u03b5\u03b3\u03c7\u03cc\u03bc\u03b5\u03bd\u03bf workflow<\/p><h3>\u039c\u03b5\u03c4\u03c1\u03ae\u03c3\u03c4\u03b5 \u03c4\u03bf \u03c0\u03c1\u03b1\u03b3\u03bc\u03b1\u03c4\u03b9\u03ba\u03cc task \u03c0\u03c1\u03b9\u03bd \u03b4\u03ce\u03c3\u03b5\u03c4\u03b5 \u03b4\u03b9\u03ba\u03b1\u03b9\u03ce\u03bc\u03b1\u03c4\u03b1 \u03c0\u03b1\u03c1\u03b1\u03b3\u03c9\u03b3\u03ae\u03c2<\/h3><p>\u0397 TWO DOTS \u03c7\u03b1\u03c1\u03c4\u03bf\u03b3\u03c1\u03b1\u03c6\u03b5\u03af \u03c4\u03b7 \u03c1\u03bf\u03ae, \u03bf\u03c1\u03af\u03b6\u03b5\u03b9 checkers, approval gates, logging \u03ba\u03b1\u03b9 fallback \u03ba\u03b1\u03b9 \u03b4\u03bf\u03ba\u03b9\u03bc\u03ac\u03b6\u03b5\u03b9 AI automations \u03c3\u03b5 e-commerce, marketing, support \u03ba\u03b1\u03b9 back-office \u03c0\u03b5\u03c1\u03b9\u03b2\u03ac\u03bb\u03bb\u03bf\u03bd\u03c4\u03b1. \u03a4\u03bf pilot \u03be\u03b5\u03ba\u03b9\u03bd\u03ac \u03bc\u03b5 \u03c0\u03b5\u03c1\u03b9\u03bf\u03c1\u03b9\u03c3\u03bc\u03ad\u03bd\u03bf scope \u03ba\u03b1\u03b9 \u03ba\u03b1\u03c4\u03b1\u03bb\u03ae\u03b3\u03b5\u03b9 \u03c3\u03b5 \u03c4\u03b5\u03ba\u03bc\u03b7\u03c1\u03b9\u03c9\u03bc\u03ad\u03bd\u03bf go\/no-go.<\/p><div class=\"td-service-cta-actions\"><a class=\"td-service-cta-button\" href=\"https:\/\/twodots.gr\/aftomatismoi-epicheiriseon-ai\/\">\u0394\u03b5\u03af\u03c4\u03b5 \u03c4\u03bf\u03c5\u03c2 \u0391\u03c5\u03c4\u03bf\u03bc\u03b1\u03c4\u03b9\u03c3\u03bc\u03bf\u03cd\u03c2 \u0395\u03c0\u03b9\u03c7\u03b5\u03b9\u03c1\u03ae\u03c3\u03b5\u03c9\u03bd &amp; AI<\/a><\/div><\/div><\/section>\n\n\n\n<section id=\"sychnes-erotiseis\" class=\"td-faq-section\"><div class=\"td-faq\"><p class=\"td-faq-heading\">Frequently Asked Questions (FAQs)<\/p><details class=\"td-faq-item\"><summary class=\"td-faq-title\">\u03a4\u03b9 \u03b5\u03af\u03bd\u03b1\u03b9 \u03c4\u03bf MAG benchmark;<\/summary><div class=\"td-faq-content\"><p>\u03a4\u03bf MAG, \u03b1\u03c0\u03cc \u03c4\u03b1 \u03b1\u03c1\u03c7\u03b9\u03ba\u03ac Multimodal Action and Guide, \u03b5\u03af\u03bd\u03b1\u03b9 \u03b5\u03c1\u03b5\u03c5\u03bd\u03b7\u03c4\u03b9\u03ba\u03cc benchmark \u03b3\u03b9\u03b1 web agents \u03c0\u03bf\u03c5 \u03bc\u03b5\u03c4\u03c1\u03ac \u03bc\u03b1\u03b6\u03af \u03c4\u03b7\u03bd \u03bf\u03bb\u03bf\u03ba\u03bb\u03ae\u03c1\u03c9\u03c3\u03b7 \u03c0\u03bf\u03bb\u03c5\u03c3\u03c4\u03b1\u03b4\u03b9\u03b1\u03ba\u03ce\u03bd \u03b5\u03c1\u03b3\u03b1\u03c3\u03b9\u03ce\u03bd \u03c3\u03b5 live websites \u03ba\u03b1\u03b9 \u03c4\u03b7 \u03b4\u03b7\u03bc\u03b9\u03bf\u03c5\u03c1\u03b3\u03af\u03b1 \u03b1\u03bd\u03b8\u03c1\u03ce\u03c0\u03b9\u03bd\u03c9\u03bd \u03bf\u03b4\u03b7\u03b3\u03b9\u03ce\u03bd \u03b3\u03b9\u03b1 \u03ba\u03ac\u03b8\u03b5 \u03b2\u03ae\u03bc\u03b1.<\/p><\/div><\/details><details class=\"td-faq-item\"><summary class=\"td-faq-title\">\u03a0\u03bf\u03b9\u03b1 \u03ae\u03c4\u03b1\u03bd \u03b7 \u03ba\u03b1\u03bb\u03cd\u03c4\u03b5\u03c1\u03b7 \u03b5\u03c0\u03af\u03b4\u03bf\u03c3\u03b7 \u03c3\u03c4\u03bf MAG;<\/summary><div class=\"td-faq-content\"><p>\u0397 \u03ba\u03b1\u03bb\u03cd\u03c4\u03b5\u03c1\u03b7 \u03b4\u03b9\u03b1\u03bc\u03cc\u03c1\u03c6\u03c9\u03c3\u03b7 \u03c4\u03b7\u03c2 \u03bc\u03b5\u03bb\u03ad\u03c4\u03b7\u03c2 \u03bf\u03bb\u03bf\u03ba\u03bb\u03ae\u03c1\u03c9\u03c3\u03b5 65 \u03b1\u03c0\u03cc \u03c4\u03b1 174 test tasks, \u03b4\u03b7\u03bb\u03b1\u03b4\u03ae 37,4%. \u03a4\u03bf \u03b1\u03c0\u03bf\u03c4\u03ad\u03bb\u03b5\u03c3\u03bc\u03b1 \u03b4\u03b5\u03af\u03c7\u03bd\u03b5\u03b9 \u03c0\u03c1\u03cc\u03bf\u03b4\u03bf, \u03b1\u03bb\u03bb\u03ac \u03cc\u03c7\u03b9 \u03b1\u03be\u03b9\u03bf\u03c0\u03b9\u03c3\u03c4\u03af\u03b1 \u03ba\u03b1\u03c4\u03ac\u03bb\u03bb\u03b7\u03bb\u03b7 \u03b3\u03b9\u03b1 \u03b1\u03bd\u03b5\u03be\u03ad\u03bb\u03b5\u03b3\u03ba\u03c4\u03b7 \u03c0\u03b1\u03c1\u03b1\u03b3\u03c9\u03b3\u03b9\u03ba\u03ae \u03c7\u03c1\u03ae\u03c3\u03b7.<\/p><\/div><\/details><details class=\"td-faq-item\"><summary class=\"td-faq-title\">\u03a4\u03b9 \u03b4\u03b9\u03b1\u03c6\u03bf\u03c1\u03ac \u03ad\u03c7\u03b5\u03b9 \u03c4\u03bf Set-of-Mark \u03b1\u03c0\u03cc \u03c4\u03b9\u03c2 \u03c3\u03c5\u03bd\u03c4\u03b5\u03c4\u03b1\u03b3\u03bc\u03ad\u03bd\u03b5\u03c2;<\/summary><div class=\"td-faq-content\"><p>\u03a3\u03c4\u03bf Set-of-Mark \u03bf agent \u03b5\u03c0\u03b9\u03bb\u03ad\u03b3\u03b5\u03b9 \u03b1\u03c1\u03b9\u03b8\u03bc\u03b7\u03bc\u03ad\u03bd\u03b7 \u03c3\u03ae\u03bc\u03b1\u03bd\u03c3\u03b7 \u03c0\u03ac\u03bd\u03c9 \u03c3\u03b5 \u03b4\u03b9\u03b1\u03b4\u03c1\u03b1\u03c3\u03c4\u03b9\u03ba\u03cc \u03c3\u03c4\u03bf\u03b9\u03c7\u03b5\u03af\u03bf. \u03a3\u03c4\u03b9\u03c2 \u03c3\u03c5\u03bd\u03c4\u03b5\u03c4\u03b1\u03b3\u03bc\u03ad\u03bd\u03b5\u03c2 \u03c3\u03c4\u03bf\u03c7\u03b5\u03cd\u03b5\u03b9 \u03b1\u03c0\u03b5\u03c5\u03b8\u03b5\u03af\u03b1\u03c2 \u03b8\u03ad\u03c3\u03b7 pixel. \u03a4\u03bf MAG \u03b2\u03c1\u03ae\u03ba\u03b5 \u03cc\u03c4\u03b9 \u03b7 \u03ba\u03b1\u03bb\u03cd\u03c4\u03b5\u03c1\u03b7 \u03b5\u03c0\u03b9\u03bb\u03bf\u03b3\u03ae \u03b5\u03be\u03b1\u03c1\u03c4\u03ac\u03c4\u03b1\u03b9 \u03b1\u03c0\u03cc \u03c4\u03bf \u03bc\u03bf\u03bd\u03c4\u03ad\u03bb\u03bf.<\/p><\/div><\/details><details class=\"td-faq-item\"><summary class=\"td-faq-title\">\u0393\u03b9\u03b1\u03c4\u03af \u03c4\u03bf MAG \u03b6\u03b7\u03c4\u03ac \u03ba\u03b1\u03b9 \u03bf\u03b4\u03b7\u03b3\u03af\u03b5\u03c2 \u03b3\u03b9\u03b1 \u03c4\u03bf\u03bd \u03c7\u03c1\u03ae\u03c3\u03c4\u03b7;<\/summary><div class=\"td-faq-content\"><p>\u0395\u03c0\u03b5\u03b9\u03b4\u03ae \u03bc\u03b9\u03b1 \u03b5\u03c0\u03b9\u03c4\u03c5\u03c7\u03b7\u03bc\u03ad\u03bd\u03b7 web \u03b4\u03b9\u03b1\u03b4\u03c1\u03bf\u03bc\u03ae \u03ad\u03c7\u03b5\u03b9 \u03bc\u03b5\u03b3\u03b1\u03bb\u03cd\u03c4\u03b5\u03c1\u03b7 \u03b1\u03be\u03af\u03b1 \u03cc\u03c4\u03b1\u03bd \u03bc\u03c0\u03bf\u03c1\u03b5\u03af \u03bd\u03b1 \u03bc\u03b5\u03c4\u03b1\u03c4\u03c1\u03b1\u03c0\u03b5\u03af \u03c3\u03b5 \u03ba\u03b1\u03c4\u03b1\u03bd\u03bf\u03b7\u03c4\u03ae, \u03b5\u03c0\u03b1\u03bd\u03b1\u03c7\u03c1\u03b7\u03c3\u03b9\u03bc\u03bf\u03c0\u03bf\u03b9\u03ae\u03c3\u03b9\u03bc\u03b7 \u03ba\u03b1\u03b8\u03bf\u03b4\u03ae\u03b3\u03b7\u03c3\u03b7. \u0397 \u03bf\u03b4\u03b7\u03b3\u03af\u03b1 \u03b2\u03b1\u03b8\u03bc\u03bf\u03bb\u03bf\u03b3\u03b5\u03af\u03c4\u03b1\u03b9 \u03c7\u03c9\u03c1\u03b9\u03c3\u03c4\u03ac \u03ba\u03b1\u03b9 \u03b4\u03b5\u03bd \u03bc\u03c0\u03bf\u03c1\u03b5\u03af \u03bd\u03b1 \u03ba\u03b1\u03bb\u03cd\u03c8\u03b5\u03b9 \u03bc\u03b9\u03b1 \u03b1\u03c0\u03bf\u03c4\u03c5\u03c7\u03b7\u03bc\u03ad\u03bd\u03b7 \u03b5\u03bd\u03ad\u03c1\u03b3\u03b5\u03b9\u03b1.<\/p><\/div><\/details><details class=\"td-faq-item\"><summary class=\"td-faq-title\">\u0395\u03af\u03bd\u03b1\u03b9 \u03b4\u03b9\u03b1\u03b8\u03ad\u03c3\u03b9\u03bc\u03bf \u03c3\u03ae\u03bc\u03b5\u03c1\u03b1 \u03c9\u03c2 \u03ad\u03c4\u03bf\u03b9\u03bc\u03bf \u03b5\u03c1\u03b3\u03b1\u03bb\u03b5\u03af\u03bf;<\/summary><div class=\"td-faq-content\"><p>\u0397 \u03c3\u03b5\u03bb\u03af\u03b4\u03b1 \u03c4\u03bf\u03c5 arXiv \u03b1\u03bd\u03b1\u03c6\u03ad\u03c1\u03b5\u03b9 \u03cc\u03c4\u03b9 benchmark, harness \u03ba\u03b1\u03b9 checkpoints \u03c0\u03c1\u03cc\u03ba\u03b5\u03b9\u03c4\u03b1\u03b9 \u03bd\u03b1 \u03b4\u03b9\u03b1\u03c4\u03b5\u03b8\u03bf\u03cd\u03bd. \u0395\u03c0\u03bf\u03bc\u03ad\u03bd\u03c9\u03c2 \u03b7 \u03c3\u03b7\u03bc\u03b5\u03c1\u03b9\u03bd\u03ae \u03c0\u03c1\u03b1\u03ba\u03c4\u03b9\u03ba\u03ae \u03b1\u03be\u03af\u03b1 \u03c4\u03bf\u03c5 MAG \u03b5\u03af\u03bd\u03b1\u03b9 \u03ba\u03c5\u03c1\u03af\u03c9\u03c2 \u03c9\u03c2 \u03c0\u03bb\u03b1\u03af\u03c3\u03b9\u03bf \u03b1\u03be\u03b9\u03bf\u03bb\u03cc\u03b3\u03b7\u03c3\u03b7\u03c2 \u03ba\u03b1\u03b9 \u03cc\u03c7\u03b9 \u03c9\u03c2 \u03ad\u03c4\u03bf\u03b9\u03bc\u03b7 \u03b5\u03bc\u03c0\u03bf\u03c1\u03b9\u03ba\u03ae \u03c0\u03bb\u03b1\u03c4\u03c6\u03cc\u03c1\u03bc\u03b1.<\/p><\/div><\/details><details class=\"td-faq-item\"><summary class=\"td-faq-title\">\u039c\u03c0\u03bf\u03c1\u03b5\u03af \u03ad\u03bd\u03b1\u03c2 web agent \u03bd\u03b1 \u03c7\u03c1\u03b7\u03c3\u03b9\u03bc\u03bf\u03c0\u03bf\u03b9\u03b7\u03b8\u03b5\u03af \u03c3\u03b5 e-commerce;<\/summary><div class=\"td-faq-content\"><p>\u039d\u03b1\u03b9, \u03b1\u03c1\u03c7\u03b9\u03ba\u03ac \u03c3\u03b5 \u03c0\u03b5\u03c1\u03b9\u03bf\u03c1\u03b9\u03c3\u03bc\u03ad\u03bd\u03b1 \u03ba\u03b1\u03b9 \u03b1\u03bd\u03b1\u03c3\u03c4\u03c1\u03ad\u03c8\u03b9\u03bc\u03b1 tasks \u03cc\u03c0\u03c9\u03c2 \u03c3\u03c5\u03bb\u03bb\u03bf\u03b3\u03ae \u03c3\u03c4\u03bf\u03b9\u03c7\u03b5\u03af\u03c9\u03bd, \u03c0\u03c1\u03bf\u03b5\u03c4\u03bf\u03b9\u03bc\u03b1\u03c3\u03af\u03b1 \u03c0\u03c1\u03bf\u03c3\u03c7\u03b5\u03b4\u03af\u03bf\u03c5 \u03ae \u03ad\u03bb\u03b5\u03b3\u03c7\u03bf\u03c2 \u03c0\u03bb\u03b7\u03c1\u03cc\u03c4\u03b7\u03c4\u03b1\u03c2. \u0391\u03bb\u03bb\u03b1\u03b3\u03ad\u03c2 \u03c3\u03b5 \u03c0\u03b1\u03c1\u03b1\u03b3\u03b3\u03b5\u03bb\u03af\u03b5\u03c2, \u03c7\u03c1\u03ae\u03bc\u03b1\u03c4\u03b1 \u03ae \u03b5\u03c0\u03b9\u03ba\u03bf\u03b9\u03bd\u03c9\u03bd\u03af\u03b1 \u03bc\u03b5 \u03c0\u03b5\u03bb\u03ac\u03c4\u03b7 \u03c7\u03c1\u03b5\u03b9\u03ac\u03b6\u03bf\u03bd\u03c4\u03b1\u03b9 \u03b1\u03bd\u03b5\u03be\u03ac\u03c1\u03c4\u03b7\u03c4\u03bf checker \u03ba\u03b1\u03b9 \u03b1\u03bd\u03b8\u03c1\u03ce\u03c0\u03b9\u03bd\u03b7 \u03ad\u03b3\u03ba\u03c1\u03b9\u03c3\u03b7.<\/p><\/div><\/details><details class=\"td-faq-item\"><summary class=\"td-faq-title\">\u03a0\u03ce\u03c2 \u03c0\u03c1\u03ad\u03c0\u03b5\u03b9 \u03bd\u03b1 \u03b5\u03bb\u03ad\u03b3\u03c7\u03b5\u03c4\u03b1\u03b9 \u03b7 \u03b5\u03c0\u03b9\u03c4\u03c5\u03c7\u03af\u03b1 \u03b5\u03bd\u03cc\u03c2 agent;<\/summary><div class=\"td-faq-content\"><p>\u039c\u03b5 \u03b1\u03bd\u03c4\u03b9\u03ba\u03b5\u03b9\u03bc\u03b5\u03bd\u03b9\u03ba\u03cc readback \u03c4\u03b7\u03c2 \u03c4\u03b5\u03bb\u03b9\u03ba\u03ae\u03c2 \u03ba\u03b1\u03c4\u03ac\u03c3\u03c4\u03b1\u03c3\u03b7\u03c2 \u03bc\u03ad\u03c3\u03c9 API, \u03b2\u03ac\u03c3\u03b7\u03c2 \u03ae \u03ba\u03b1\u03bd\u03cc\u03bd\u03b1 \u03cc\u03c0\u03bf\u03c5 \u03b5\u03af\u03bd\u03b1\u03b9 \u03b4\u03c5\u03bd\u03b1\u03c4\u03cc, \u03bc\u03b1\u03b6\u03af \u03bc\u03b5 action log \u03ba\u03b1\u03b9 screenshot. \u03a4\u03bf \u03bc\u03ae\u03bd\u03c5\u03bc\u03b1 \u00ab\u03bf\u03bb\u03bf\u03ba\u03bb\u03b7\u03c1\u03ce\u03b8\u03b7\u03ba\u03b5\u00bb \u03c4\u03bf\u03c5 \u03af\u03b4\u03b9\u03bf\u03c5 \u03c4\u03bf\u03c5 agent \u03b4\u03b5\u03bd \u03b1\u03c1\u03ba\u03b5\u03af \u03c9\u03c2 \u03b1\u03c0\u03cc\u03b4\u03b5\u03b9\u03be\u03b7.<\/p><\/div><\/details><details class=\"td-faq-item\"><summary class=\"td-faq-title\">\u03a0\u03bf\u03b9\u03bf \u03b5\u03af\u03bd\u03b1\u03b9 \u03c4\u03bf \u03b2\u03b1\u03c3\u03b9\u03ba\u03cc \u03c3\u03c5\u03bc\u03c0\u03ad\u03c1\u03b1\u03c3\u03bc\u03b1 \u03b3\u03b9\u03b1 \u03bc\u03b9\u03b1 \u03b5\u03c0\u03b9\u03c7\u03b5\u03af\u03c1\u03b7\u03c3\u03b7;<\/summary><div class=\"td-faq-content\"><p>\u039d\u03b1 \u03b1\u03be\u03b9\u03bf\u03bb\u03bf\u03b3\u03b5\u03af web agents \u03c3\u03c4\u03b1 \u03b4\u03b9\u03ba\u03ac \u03c4\u03b7\u03c2 tasks, \u03b1\u03bd\u03ac \u03bc\u03ae\u03ba\u03bf\u03c2 \u03c1\u03bf\u03ae\u03c2 \u03ba\u03b1\u03b9 \u03b5\u03c0\u03af\u03c0\u03b5\u03b4\u03bf \u03c1\u03af\u03c3\u03ba\u03bf\u03c5, \u03bc\u03b5 \u03c0\u03b5\u03c1\u03b9\u03bf\u03c1\u03b9\u03c3\u03bc\u03ad\u03bd\u03b1 \u03b4\u03b9\u03ba\u03b1\u03b9\u03ce\u03bc\u03b1\u03c4\u03b1, shadow mode, approval gates \u03ba\u03b1\u03b9 \u03c3\u03b1\u03c6\u03ae fallback \u03c0\u03c1\u03b9\u03bd \u03b1\u03c0\u03cc \u03bf\u03c0\u03bf\u03b9\u03b1\u03b4\u03ae\u03c0\u03bf\u03c4\u03b5 \u03c0\u03b1\u03c1\u03b1\u03b3\u03c9\u03b3\u03b9\u03ba\u03ae \u03b1\u03c5\u03c4\u03bf\u03bd\u03bf\u03bc\u03af\u03b1.<\/p><\/div><\/details><\/div><\/section>\n\n\n\n<div class=\"td-source-list\"><p id=\"piges\" class=\"td-source-list-title\">Sources<\/p><ul><li><a href=\"https:\/\/arxiv.org\/abs\/2607.10079\" target=\"_blank\" rel=\"noopener\">Gan et al. \u2014 MAG: A Web-Agent Benchmark and Harness for Multimodal Action and Guide Generation, arXiv v3 (2026)<\/a><\/li><li><a href=\"https:\/\/arxiv.org\/abs\/2307.13854\" target=\"_blank\" rel=\"noopener\">Zhou et al. \u2014 WebArena: A Realistic Web Environment for Building Autonomous Agents<\/a><\/li><li><a href=\"https:\/\/www.nist.gov\/publications\/artificial-intelligence-risk-management-framework-generative-artificial-intelligence\" target=\"_blank\" rel=\"noopener\">NIST AI 600-1 \u2014 Generative Artificial Intelligence Profile<\/a><\/li><\/ul><\/div>","protected":false},"excerpt":{"rendered":"<p>\u03a4\u03bf MAG benchmark \u03bc\u03b5\u03c4\u03c1\u03ac \u03b1\u03bd \u03bf\u03b9 AI agents \u03bf\u03bb\u03bf\u03ba\u03bb\u03b7\u03c1\u03ce\u03bd\u03bf\u03c5\u03bd web \u03b5\u03c1\u03b3\u03b1\u03c3\u03af\u03b5\u03c2 \u03ba\u03b1\u03b9 \u03b3\u03c1\u03ac\u03c6\u03bf\u03c5\u03bd \u03c7\u03c1\u03ae\u03c3\u03b9\u03bc\u03b5\u03c2 \u03bf\u03b4\u03b7\u03b3\u03af\u03b5\u03c2. \u0394\u03b5\u03af\u03c4\u03b5 \u03b5\u03c0\u03b9\u03b4\u03cc\u03c3\u03b5\u03b9\u03c2, \u03cc\u03c1\u03b9\u03b1 \u03ba\u03b1\u03b9 \u03b1\u03c3\u03c6\u03b1\u03bb\u03ae \u03b5\u03c0\u03b9\u03c7\u03b5\u03b9\u03c1\u03b7\u03bc\u03b1\u03c4\u03b9\u03ba\u03ae \u03c7\u03c1\u03ae\u03c3\u03b7.<\/p>","protected":false},"author":1,"featured_media":86155,"comment_status":"open","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"_gspb_post_css":"","footnotes":""},"categories":[199],"tags":[6506,3854,18468,18467,18466],"class_list":["post-85468","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-techniti-noimosyni","tag-ai-agents","tag-customer-experience","tag-digital-adoption","tag-multimodal-ai","tag-web-automation"],"blocksy_meta":[],"_links":{"self":[{"href":"https:\/\/twodots.gr\/en\/wp-json\/wp\/v2\/posts\/85468","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/twodots.gr\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/twodots.gr\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/twodots.gr\/en\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/twodots.gr\/en\/wp-json\/wp\/v2\/comments?post=85468"}],"version-history":[{"count":1,"href":"https:\/\/twodots.gr\/en\/wp-json\/wp\/v2\/posts\/85468\/revisions"}],"predecessor-version":[{"id":86156,"href":"https:\/\/twodots.gr\/en\/wp-json\/wp\/v2\/posts\/85468\/revisions\/86156"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/twodots.gr\/en\/wp-json\/wp\/v2\/media\/86155"}],"wp:attachment":[{"href":"https:\/\/twodots.gr\/en\/wp-json\/wp\/v2\/media?parent=85468"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/twodots.gr\/en\/wp-json\/wp\/v2\/categories?post=85468"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/twodots.gr\/en\/wp-json\/wp\/v2\/tags?post=85468"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}