{"url":"https://tokumini.hatenadiary.com/entry/2021/07/10/235900","blog_url":"https://tokumini.hatenadiary.com/","provider_name":"Hatena Blog","provider_url":"https://hatena.blog","version":"1.0","width":"100%","height":"190","author_url":"https://blog.hatena.ne.jp/tokumini/","author_name":"tokumini","blog_title":"\u65e5\u8a18","type":"rich","image_url":null,"categories":[],"published":"2021-07-10 23:59:00","title":"20210710","html":"<iframe src=\"https://hatenablog-parts.com/embed?url=https%3A%2F%2Ftokumini.hatenadiary.com%2Fentry%2F2021%2F07%2F10%2F235900\" title=\"20210710 - \u65e5\u8a18\" class=\"embed-card embed-blogcard\" scrolling=\"no\" frameborder=\"0\" style=\"display: block; width: 100%; height: 190px; max-width: 500px; margin: 10px 0px;\"></iframe>","description":"My lab is building neural nets that emulate human \u201cthinking\u201d processes. They increase their intelligence after training by increasing their compute budget. By doing so, a net trained only on \u201ceasy\u201d chess puzzles can solve \u201chard\u201d chess puzzles without having ever seen one\u2026 pic.twitter.com/Wh197CzXkC\u2014\u2026"}