[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"product:meta-are":3},{"id":4,"name":5,"slug":6,"slogan":7,"description":8,"organization":9,"images":10,"avatar_url":11,"image_url":12,"website_url":13,"access_links":14,"status":19,"build_by":20,"categories":21,"stats":25,"upload_relationship":28,"create_at":29,"update_at":30,"owner_id":26,"upload_user_id":31,"reviewer_id":31,"video_info":32,"prerelease_cover_url":33,"scheduled_publish_at":32,"user_scheduled":34,"emoji_reacts":35,"tag":32,"pass_review":32,"user_star":34},671,"Meta ARE","meta-are","用于 Agent 评估的动态仿真研究平台。","{\"type\": \"doc\", \"content\": [{\"type\": \"paragraph\", \"content\": [{\"text\": \"Meta Agents Research Environments (ARE) 是一个动态仿真研究平台，用于在复杂的多步骤任务上训练和评估 AI 智能体，这些任务反映了现实世界的挑战。与静态基准测试不同，ARE 创建了不断演化的环境，智能体必须在新信息出现和条件随时间变化时调整其策略。特别地，ARE 运行 Gaia2 基准测试，这是 Gaia 的后续版本，评估更广泛的智能体能力。\", \"type\": \"text\"}]}, {\"type\": \"paragraph\"}, {\"type\": \"paragraph\", \"content\": [{\"text\": \"【ARE 能做什么？】\", \"type\": \"text\"}]}, {\"type\": \"paragraph\", \"content\": [{\"text\": \"1. 动态仿真\", \"type\": \"text\"}]}, {\"type\": \"paragraph\", \"content\": [{\"text\": \"创建在几分钟、几小时或几天内演化的现实场景——模拟需要持续推理和适应的复杂工作流程。\", \"type\": \"text\"}]}, {\"type\": \"paragraph\", \"content\": [{\"text\": \"2. 智能体评估\", \"type\": \"text\"}]}, {\"type\": \"paragraph\", \"content\": [{\"text\": \"使用综合基准测试工具在多步骤任务上测试 AI 智能体，包括跨 10 个宇宙的 800 个场景的 Gaia2 基准测试。\", \"type\": \"text\"}]}, {\"type\": \"paragraph\", \"content\": [{\"text\": \"3. 交互式应用\", \"type\": \"text\"}]}, {\"type\": \"paragraph\", \"content\": [{\"text\": \"智能体与现实的应用程序交互，如电子邮件、日历、文件系统和消息传递——每个都具有特定领域的数据和行为。\", \"type\": \"text\"}]}, {\"type\": \"paragraph\", \"content\": [{\"text\": \"4. 研究与基准测试\", \"type\": \"text\"}]}, {\"type\": \"paragraph\", \"content\": [{\"text\": \"通过并行执行、多模型支持和为研究社区自动收集结果进行系统性评估。\", \"type\": \"text\"}]}, {\"type\": \"paragraph\"}]}","Meta","https:\u002F\u002Fwatcha.tos-cn-beijing.volces.com\u002Fprod\u002Fuser\u002Fuploads\u002F2_1759737842_da5f5539-0474-4286-a434-9177a4def199.png;https:\u002F\u002Fwatcha.tos-cn-beijing.volces.com\u002Fprod\u002Fuser\u002Fuploads\u002F2_1759737842_0b586bd4-40bc-401a-a104-5fa3bb987f33.png;https:\u002F\u002Fwatcha.tos-cn-beijing.volces.com\u002Fprod\u002Fuser\u002Fuploads\u002F2_1759737842_fab9a5a1-9792-4b68-add5-c769c91e85fa.png","https:\u002F\u002Fwatcha.tos-cn-beijing.volces.com\u002Fprod\u002Fuser\u002Fuploads\u002F2_1759737808_8eefeb17-ac5e-483f-bcdd-3f2080872363.png","https:\u002F\u002Fwatcha.tos-cn-beijing.volces.com\u002Fprod\u002Fuser\u002Fuploads\u002F2_1759737863_0d572043-23e5-4b54-a954-69e797cd47ef.png","https:\u002F\u002Ffacebookresearch.github.io\u002Fmeta-agents-research-environments\u002F",{"items":15},[16],{"platform":17,"url":13,"is_primary":18},"WEB",true,"PUBLISHED",[],[22],{"id":23,"name":24},12,"其他类型",{"upvotes":26,"stars":26,"review_count":26,"reply_count":26,"update_at":27},0,"2026-09-09T08:02:22.457Z","THIRD_PARTY","2025-10-06T08:04:36.171Z","2025-10-06T08:06:08.054Z",2,null,"",false,{"reacts":36},[]]