{"id":3742,"date":"2023-09-07T05:29:00","date_gmt":"2023-09-06T20:29:00","guid":{"rendered":"https:\/\/devneko.jp\/wordpress\/?p=3742"},"modified":"2023-09-07T05:29:00","modified_gmt":"2023-09-06T20:29:00","slug":"mental-llm","status":"publish","type":"post","link":"https:\/\/devneko.jp\/wordpress\/?p=3742","title":{"rendered":"Mental-LLM"},"content":{"rendered":"\n<ul class=\"wp-block-list\">\n<li><strong>Mental-LLM: Leveraging Large Language Models for Mental Health Prediction via Online Text Data&nbsp;<\/strong>[38.9]<br>\u30aa\u30f3\u30e9\u30a4\u30f3\u30c6\u30ad\u30b9\u30c8\u30c7\u30fc\u30bf\u3092\u7528\u3044\u3066,\u69d8\u3005\u306a\u30e1\u30f3\u30bf\u30eb\u30d8\u30eb\u30b9\u4e88\u6e2c\u30bf\u30b9\u30af\u306b\u304a\u3051\u308b\u591a\u8a00\u8a9e\u30e2\u30c7\u30eb(LLM)\u306e\u7dcf\u5408\u8a55\u4fa1\u3092\u884c\u3063\u305f\u3002 \u305d\u306e\u7d50\u679c\u3001\u30bc\u30ed\u30b7\u30e7\u30c3\u30c8\u30d7\u30ed\u30f3\u30d7\u30c8\u3001\u5c11\u6570\u30b7\u30e7\u30c3\u30c8\u30d7\u30ed\u30f3\u30d7\u30c8\u3001\u547d\u4ee4\u5fae\u8abf\u6574\u306b\u3088\u308bLLM\u306e\u6709\u671b\u306a\u6027\u80fd\u304c\u793a\u3055\u308c\u305f\u3002 \u6211\u3005\u306e\u6700\u3082\u7cbe\u5de7\u306a\u30e2\u30c7\u30eb\u3067\u3042\u308bMental-Alpaca\u3068Mental-FLAN-T5\u306f\u3001\u30d0\u30e9\u30f3\u30b9\u306e\u53d6\u308c\u305f\u7cbe\u5ea6\u3067GPT-3.5\u309210.9%\u4e0a\u56de\u308a\u3001GPT-4(250\u500d\u3001150\u500d)\u30924.8%\u4e0a\u56de\u308a\u307e\u3057\u305f\u3002&nbsp;<br><a href=\"http:\/\/arxiv.org\/abs\/2307.14385v2\">\u8ad6\u6587<\/a>&nbsp;&nbsp;<a href=\"https:\/\/fugumt.com\/fugumt\/paper_check\/2307.14385v2\">\u53c2\u8003\u8a33\uff08\u30e1\u30bf\u30c7\u30fc\u30bf\uff09<\/a>&nbsp; &nbsp;(Wed, 16 Aug 2023 06:04:48 GMT)<\/li>\n\n\n\n<li>\u30e1\u30f3\u30bf\u30eb\u30d8\u30eb\u30b9\u4e88\u6e2c\u30bf\u30b9\u30af\u3078\u306eLLM\u6d3b\u7528\u306b\u95a2\u3059\u308b\u5831\u544a\u3002zero shot, few shot, instruction finetuning\u3068\u3044\u3063\u305f\u69d8\u3005\u306a\u65b9\u6cd5\uff06Alpaca, FLAN, GPT-3.5, GPT-4\u306a\u3069\u69d8\u3005\u306a\u30e2\u30c7\u30eb\u3067\u8a55\u4fa1\u3055\u308c\u3066\u304a\u308a\u8208\u5473\u6df1\u3044\u3002 <\/li>\n\n\n\n<li>\u300cInstruction finetuning on multiple mental health datasets can significantly boost the performance of LLMs on various mental health prediction tasks.\u300d\u3084\u300cAlthough task-solving-focused LLMs may have better performance in the zero-shot setting for mental health prediction tasks, dialogue-focused LLMs have a stronger capability of learning from human natural language and can improve more significantly after finetuning.\u300d\u306a\u3069\u8208\u5473\u6df1\u3044\u7d50\u679c\u3068\u306a\u3063\u3066\u3044\u308b\u3002\u30e1\u30f3\u30bf\u30eb\u30d8\u30eb\u30b9\u3068\u3044\u3046\u30c9\u30e1\u30a4\u30f3\u306b\u4f9d\u5b58\u3057\u305f\u8a71\u306a\u306e\u304b\u4e00\u822c\u7684\u306a\u50be\u5411\u306a\u306e\u304b\u306f\u3088\u304f\u308f\u304b\u3089\u306a\u3044\u304c\u3001\u69d8\u3005\u306a\u30a2\u30d7\u30ed\u30fc\u30c1\u306e\u7d50\u679c\u3092\u6bd4\u8f03\u3059\u308b\u306e\u306f\u91cd\u8981\u3067\u3042\u308b\u3068\u306e\u611f\u60f3\u3002<\/li>\n<\/ul>\n","protected":false},"excerpt":{"rendered":"","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[2],"tags":[223,558],"class_list":["post-3742","post","type-post","status-publish","format-standard","hentry","category-arxiv","tag-llm","tag-558"],"_links":{"self":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/3742","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=3742"}],"version-history":[{"count":0,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/3742\/revisions"}],"wp:attachment":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=3742"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=3742"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=3742"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}