{"id":5775,"date":"2024-11-21T05:50:00","date_gmt":"2024-11-20T20:50:00","guid":{"rendered":"https:\/\/devneko.jp\/wordpress\/?p=5775"},"modified":"2024-11-21T05:50:00","modified_gmt":"2024-11-20T20:50:00","slug":"decoprompt","status":"publish","type":"post","link":"https:\/\/devneko.jp\/wordpress\/?p=5775","title":{"rendered":"DecoPrompt"},"content":{"rendered":"\n<ul class=\"wp-block-list\">\n<li><strong>DecoPrompt : Decoding Prompts Reduces Hallucinations when Large Language Models Meet False Premises\u00a0<\/strong>[28.7]<br>\u5e7b\u899a\u3092\u7de9\u548c\u3059\u308b\u65b0\u3057\u3044\u30d7\u30ed\u30f3\u30d7\u30c8\u30a2\u30eb\u30b4\u30ea\u30ba\u30e0DecoPrompt\u3092\u63d0\u6848\u3059\u308b\u3002 DecoPrompt \u306f LLM \u3092\u5229\u7528\u3057\u3066\u507d\u524d\u63d0\u306e\u30d7\u30ed\u30f3\u30d7\u30c8\u3092 &#8220;\u30c7\u30b3\u30fc\u30c9&#8221; \u3059\u308b\u3002 2\u3064\u306e\u30c7\u30fc\u30bf\u30bb\u30c3\u30c8\u3067\u5b9f\u9a13\u3092\u884c\u3044\u3001DecoPrompt\u306f\u7570\u306a\u308bLLM\u304b\u3089\u51fa\u529b\u3055\u308c\u305f\u5e7b\u899a\u3092\u52b9\u679c\u7684\u306b\u4f4e\u6e1b\u3067\u304d\u308b\u3053\u3068\u3092\u793a\u3057\u305f\u3002<br><a href=\"http:\/\/arxiv.org\/abs\/2411.07457v1\">\u8ad6\u6587<\/a>\u00a0\u00a0<a href=\"https:\/\/fugumt.com\/fugumt\/paper_check\/2411.07457v1\">\u53c2\u8003\u8a33\uff08\u30e1\u30bf\u30c7\u30fc\u30bf\uff09<\/a>\u00a0 \u00a0(Tue, 12 Nov 2024 00:48:01 GMT)<\/li>\n\n\n\n<li>\u300cInspired by the observation that entropy of the false-premise prompt is closely related to its likelihood to elicit hallucination generation, we propose a new prompting algorithm, named DecoPrompt, to mitigate hallucination.\u300d\u3092\u3046\u3051\u3066\u300c1) first paraphrases the user\u2019s prompt to obtain several semantically similar candidates, then 2) decodes them with the LLM, and 3) selects the lowest-entropy candidate as the new prompt.\u300d\u3068\u3044\u3046\u624b\u6cd5\u306e\u63d0\u6848\u3002\u30b7\u30f3\u30d7\u30eb\u306a\u624b\u6cd5\u306b\u898b\u3048\u308b\u304c\u3001\u52b9\u679c\u304c\u3042\u308b\u306e\u306f\u8208\u5473\u6df1\u3044\u3002<\/li>\n\n\n\n<li>\u30ea\u30dd\u30b8\u30c8\u30ea\u306f<a href=\"https:\/\/github.com\/xunannancy\/DecoPrompt\">GitHub &#8211; xunannancy\/DecoPrompt: Code for paper DecoPrompt : Decoding Prompts Reduces Hallucinations when Large Language Models Meet False Premises<\/a><\/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":[184],"class_list":["post-5775","post","type-post","status-publish","format-standard","hentry","category-arxiv","tag-hallucination"],"_links":{"self":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/5775","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=5775"}],"version-history":[{"count":0,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/5775\/revisions"}],"wp:attachment":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=5775"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=5775"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=5775"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}