https://wuulong.github.io/wuulong-notes-blog/tags/tw-med-db/ Recent content in Tw-Med-Db on 哈爸筆記 https://wuulong.github.io/wuulong-notes-blog/%3Clink%20or%20path%20of%20image%20for%20opengraph,%20twitter-cards%3E https://wuulong.github.io/wuulong-notes-blog/%3Clink%20or%20path%20of%20image%20for%20opengraph,%20twitter-cards%3E Hugo -- 0.152.2 zh-tw Sat, 22 Aug 2026 05:10:00 +0800 https://wuulong.github.io/wuulong-notes-blog/posts/20260822_releasing_tw_med_db_open_source_engine/ Sat, 22 Aug 2026 05:10:00 +0800 https://wuulong.github.io/wuulong-notes-blog/posts/20260822_releasing_tw_med_db_open_source_engine/ <p>在推動 LLM 與 AI Agent 落地臨床或照護情境時，大家最常遇到、也最頭痛的問題，永遠不是模型不夠聰明，而是**「資料太散、太亂、又太容易讓 LLM 產生致命幻覺 (Hallucination)」**。</p>