ã¯ããã«Sakana AIããã®The AI Scientistãçé ã«ãLLM-Based Agentã®ç¤¾ä¼æ´»ç¨ã¯ä»å¾æ¡å¤§ãã¦ãã¨èãã¦ãã¾ããã¾ããAgentã§æ¨è«âçµæãåæãã¼ã¿ã¨ãã¦Tuningã«å©ç¨ã¿ãããªæµãã§ãLLMèªä½ã®çºå±ã«ãAgentã¯æ´»ç¨ããã¦ããã¨èãã¾ãã ããããã¢ããã¼ã·ã§ã³ã«æè¿LLM-Based Agentã®åå¼·ããã¦ããã試ãã«ä½ã£ããã®ãä¸é¨Noteã§å ±æã§ãããã¨æã£ã¦ãã¾ãã ä»åã¯ãLLM-Based Agentã®åºæ¬çãªã¢ã¼ããã¯ãã£ã§ããReflection Agentãé¡æã«ãSNSæ稿ã³ã³ãã³ããææ¡ãã¦ãããAgentãä½æãã¦ããããã¨æãã¾ãã LangGraphã§ã®åºæ¬çãªAgent Workflowã®æ§ç¯ã«ã¤ãã¦ã¯ä»¥åã®è¨äºããåç §ãã ããã 1. Reflection Agentã¨ã¯Reflection Agentã§ã¯ä»¥ä¸ã®
{{#tags}}- {{label}}
{{/tags}}