{"id":1236,"date":"2025-12-31T16:54:55","date_gmt":"2025-12-31T16:54:55","guid":{"rendered":"https:\/\/houshyar24.ir\/blog\/?p=1236"},"modified":"2025-12-31T16:54:55","modified_gmt":"2025-12-31T16:54:55","slug":"reinforcement-learning-in-llms","status":"publish","type":"post","link":"https:\/\/houshyar24.ir\/blog\/reinforcement-learning-in-llms\/","title":{"rendered":"\u00a0\u06cc\u0627\u062f\u06af\u06cc\u0631\u06cc \u062a\u0642\u0648\u06cc\u062a\u06cc \u062f\u0631 \u0645\u062f\u0644\u200c\u0647\u0627\u06cc \u0632\u0628\u0627\u0646\u06cc: \u0686\u0631\u0627 RL \u0642\u0644\u0628 \u062a\u067e\u0646\u062f\u0647 LLM\u0647\u0627\u06cc \u0645\u062f\u0631\u0646 \u0627\u0633\u062a\u061f"},"content":{"rendered":"<p data-start=\"199\" data-end=\"580\">\u0645\u062f\u0644\u200c\u0647\u0627\u06cc \u0632\u0628\u0627\u0646\u06cc \u0628\u0632\u0631\u06af (Large Language Models \u06cc\u0627 LLMs) \u062f\u0631 \u0633\u0627\u0644\u200c\u0647\u0627\u06cc \u0627\u062e\u06cc\u0631 \u067e\u06cc\u0634\u0631\u0641\u062a \u0686\u0634\u0645\u06af\u06cc\u0631\u06cc \u062f\u0627\u0634\u062a\u0647\u200c\u0627\u0646\u062f \u0648 \u0628\u0647 \u0627\u0628\u0632\u0627\u0631\u0647\u0627\u06cc\u06cc \u0642\u062f\u0631\u062a\u0645\u0646\u062f \u0628\u0631\u0627\u06cc \u062a\u0648\u0644\u06cc\u062f \u0645\u062a\u0646\u060c \u067e\u0627\u0633\u062e \u0628\u0647 \u067e\u0631\u0633\u0634\u200c\u0647\u0627 \u0648 \u062a\u0639\u0627\u0645\u0644 \u0632\u0628\u0627\u0646\u06cc \u062a\u0628\u062f\u06cc\u0644 \u0634\u062f\u0647\u200c\u0627\u0646\u062f. \u0628\u0627 \u0627\u06cc\u0646 \u062d\u0627\u0644\u060c \u0622\u0645\u0648\u0632\u0634 \u0627\u0648\u0644\u06cc\u0647 \u0627\u06cc\u0646 \u0645\u062f\u0644\u200c\u0647\u0627 \u0628\u0647\u200c\u062a\u0646\u0647\u0627\u06cc\u06cc \u0628\u0631\u0627\u06cc \u0631\u0633\u06cc\u062f\u0646 \u0628\u0647 \u0631\u0641\u062a\u0627\u0631\u0647\u0627\u06cc \u0645\u0637\u0644\u0648\u0628\u060c \u0627\u06cc\u0645\u0646 \u0648 \u0647\u0645\u200c\u0631\u0627\u0633\u062a\u0627 \u0628\u0627 \u0646\u06cc\u0627\u0632\u0647\u0627\u06cc \u0627\u0646\u0633\u0627\u0646\u06cc \u06a9\u0627\u0641\u06cc \u0646\u06cc\u0633\u062a.<br data-start=\"494\" data-end=\"497\" \/>\u0627\u06cc\u0646\u062c\u0627\u0633\u062a \u06a9\u0647 <strong data-start=\"508\" data-end=\"557\">\u06cc\u0627\u062f\u06af\u06cc\u0631\u06cc \u062a\u0642\u0648\u06cc\u062a\u06cc (Reinforcement Learning \u06cc\u0627 RL)<\/strong> \u0646\u0642\u0634 \u06a9\u0644\u06cc\u062f\u06cc \u067e\u06cc\u062f\u0627 \u0645\u06cc\u200c\u06a9\u0646\u062f.<\/p>\n<p data-start=\"582\" data-end=\"738\">\u062f\u0631 \u0627\u06cc\u0646 \u0628\u0644\u0627\u06af \u067e\u0633\u062a\u060c \u0628\u0647\u200c\u0635\u0648\u0631\u062a <strong data-start=\"607\" data-end=\"623\">\u06a9\u0644\u06cc \u0627\u0645\u0627 \u0639\u0644\u0645\u06cc<\/strong> \u0628\u0631\u0631\u0633\u06cc \u0645\u06cc\u200c\u06a9\u0646\u06cc\u0645 \u06a9\u0647 RL \u0686\u06cc\u0633\u062a\u060c \u0686\u0631\u0627 \u062f\u0631 \u0645\u062f\u0644\u200c\u0647\u0627\u06cc \u0632\u0628\u0627\u0646\u06cc \u0627\u0647\u0645\u06cc\u062a \u062f\u0627\u0631\u062f \u0648 \u0686\u06af\u0648\u0646\u0647 \u0628\u0647 \u0628\u0647\u0628\u0648\u062f \u06a9\u06cc\u0641\u06cc\u062a \u0648 \u0647\u0645\u200c\u0633\u0648\u06cc\u06cc \u062e\u0631\u0648\u062c\u06cc LLM\u0647\u0627 \u06a9\u0645\u06a9 \u0645\u06cc\u200c\u06a9\u0646\u062f.<\/p>\n<hr \/>\n<h2>\u00a0\u06cc\u0627\u062f\u06af\u06cc\u0631\u06cc \u062a\u0642\u0648\u06cc\u062a\u06cc (Reinforcement Learning) \u0628\u0647 \u0632\u0628\u0627\u0646 \u0633\u0627\u062f\u0647<\/h2>\n<p>\u06cc\u0627\u062f\u06af\u06cc\u0631\u06cc \u062a\u0642\u0648\u06cc\u062a\u06cc (Reinforcement Learning \u06cc\u0627 RL) \u06cc\u06a9\u06cc \u0627\u0632 \u067e\u0627\u0631\u0627\u062f\u0627\u06cc\u0645\u200c\u0647\u0627\u06cc \u0627\u0635\u0644\u06cc \u06cc\u0627\u062f\u06af\u06cc\u0631\u06cc \u0645\u0627\u0634\u06cc\u0646 \u0627\u0633\u062a \u06a9\u0647 \u062a\u0645\u0631\u06a9\u0632 \u0622\u0646 \u0628\u0631 <strong>\u062a\u0635\u0645\u06cc\u0645\u200c\u06af\u06cc\u0631\u06cc \u0645\u062a\u0648\u0627\u0644\u06cc<\/strong> \u062f\u0631 \u06cc\u06a9 \u0645\u062d\u06cc\u0637 \u067e\u0648\u06cc\u0627 \u0627\u0633\u062a.<br \/>\n\u062f\u0631 RL\u060c \u06cc\u06a9 \u0639\u0627\u0645\u0644 (Agent) \u0628\u0627 \u0627\u0646\u062c\u0627\u0645 \u0639\u0645\u0644\u200c\u0647\u0627\u06cc\u06cc \u062f\u0631 \u0645\u062d\u06cc\u0637 \u0648 \u062f\u0631\u06cc\u0627\u0641\u062a \u0628\u0627\u0632\u062e\u0648\u0631\u062f \u0639\u062f\u062f\u06cc (\u067e\u0627\u062f\u0627\u0634 \u06cc\u0627 \u062c\u0631\u06cc\u0645\u0647)\u060c \u0628\u0647\u200c\u0645\u0631\u0648\u0631 \u06cc\u0627\u062f \u0645\u06cc\u200c\u06af\u06cc\u0631\u062f \u06a9\u0647 \u0686\u06af\u0648\u0646\u0647 \u0631\u0641\u062a\u0627\u0631 \u062e\u0648\u062f \u0631\u0627 \u0628\u0647\u06cc\u0646\u0647 \u06a9\u0646\u062f.<\/p>\n<p>\u0628\u0631\u062e\u0644\u0627\u0641 \u06cc\u0627\u062f\u06af\u06cc\u0631\u06cc \u0646\u0638\u0627\u0631\u062a\u200c\u0634\u062f\u0647 \u06a9\u0647 \u067e\u0627\u0633\u062e \u062f\u0631\u0633\u062a \u0627\u0632 \u0642\u0628\u0644 \u0645\u0634\u062e\u0635 \u0627\u0633\u062a\u060c \u062f\u0631 \u06cc\u0627\u062f\u06af\u06cc\u0631\u06cc \u062a\u0642\u0648\u06cc\u062a\u06cc <strong>\u0639\u0627\u0645\u0644 \u0628\u0627\u06cc\u062f \u0628\u0627 \u0622\u0632\u0645\u0648\u0646\u200c\u0648\u062e\u0637\u0627 \u06cc\u0627\u062f \u0628\u06af\u06cc\u0631\u062f<\/strong> \u0686\u0647 \u062a\u0635\u0645\u06cc\u0645\u200c\u0647\u0627\u06cc\u06cc \u062f\u0631 \u0628\u0644\u0646\u062f\u0645\u062f\u062a \u0628\u0647\u062a\u0631\u06cc\u0646 \u0646\u062a\u06cc\u062c\u0647 \u0631\u0627 \u0645\u06cc\u200c\u062f\u0647\u0646\u062f.<\/p>\n<h3>\u0627\u062c\u0632\u0627\u06cc \u0627\u0635\u0644\u06cc \u06cc\u0627\u062f\u06af\u06cc\u0631\u06cc \u062a\u0642\u0648\u06cc\u062a\u06cc<\/h3>\n<p>\u0647\u0631 \u0633\u06cc\u0633\u062a\u0645 RL \u0627\u0632 \u0686\u0646\u062f \u0645\u0624\u0644\u0641\u0647\u200c\u06cc \u06a9\u0644\u06cc\u062f\u06cc \u062a\u0634\u06a9\u06cc\u0644 \u0634\u062f\u0647 \u0627\u0633\u062a:<\/p>\n<table style=\"height: 111px\" width=\"1151\">\n<thead>\n<tr>\n<th>\u0645\u0624\u0644\u0641\u0647<\/th>\n<th>\u062a\u0648\u0636\u06cc\u062d<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td style=\"text-align: center\"><strong>Agent (\u0639\u0627\u0645\u0644)<\/strong><\/td>\n<td style=\"text-align: center\">\u0645\u0648\u062c\u0648\u062f\u06cc\u062a\u06cc \u06a9\u0647 \u062a\u0635\u0645\u06cc\u0645 \u0645\u06cc\u200c\u06af\u06cc\u0631\u062f \u0648 \u0639\u0645\u0644 \u0627\u0646\u062c\u0627\u0645 \u0645\u06cc\u200c\u062f\u0647\u062f<\/td>\n<\/tr>\n<tr>\n<td style=\"text-align: center\"><strong>Environment (\u0645\u062d\u06cc\u0637)<\/strong><\/td>\n<td style=\"text-align: center\">\u0633\u06cc\u0633\u062a\u0645\u06cc \u06a9\u0647 \u0639\u0627\u0645\u0644 \u0628\u0627 \u0622\u0646 \u062a\u0639\u0627\u0645\u0644 \u062f\u0627\u0631\u062f<\/td>\n<\/tr>\n<tr>\n<td style=\"text-align: center\"><strong>Action (\u0639\u0645\u0644)<\/strong><\/td>\n<td style=\"text-align: center\">\u062a\u0635\u0645\u06cc\u0645 \u06cc\u0627 \u0627\u0642\u062f\u0627\u0645\u06cc \u06a9\u0647 \u0639\u0627\u0645\u0644 \u0627\u0646\u062c\u0627\u0645 \u0645\u06cc\u200c\u062f\u0647\u062f<\/td>\n<\/tr>\n<tr>\n<td style=\"text-align: center\"><strong>Reward (\u067e\u0627\u062f\u0627\u0634)<\/strong><\/td>\n<td style=\"text-align: center\">\u0628\u0627\u0632\u062e\u0648\u0631\u062f \u0639\u062f\u062f\u06cc \u0628\u0631\u0627\u06cc \u0627\u0631\u0632\u06cc\u0627\u0628\u06cc \u06a9\u06cc\u0641\u06cc\u062a \u0639\u0645\u0644<\/td>\n<\/tr>\n<tr>\n<td style=\"text-align: center\"><strong>Policy (\u0633\u06cc\u0627\u0633\u062a)<\/strong><\/td>\n<td style=\"text-align: center\">\u0627\u0633\u062a\u0631\u0627\u062a\u0698\u06cc \u0639\u0627\u0645\u0644 \u0628\u0631\u0627\u06cc \u0627\u0646\u062a\u062e\u0627\u0628 \u0639\u0645\u0644\u200c\u0647\u0627<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>\u0647\u062f\u0641 \u0639\u0627\u0645\u0644 \u0627\u06cc\u0646 \u0627\u0633\u062a \u06a9\u0647 <strong>\u0633\u06cc\u0627\u0633\u062a\u06cc (Policy)<\/strong> \u06cc\u0627\u062f \u0628\u06af\u06cc\u0631\u062f \u06a9\u0647 \u067e\u0627\u062f\u0627\u0634 \u062a\u062c\u0645\u0639\u06cc \u0631\u0627 \u062f\u0631 \u0637\u0648\u0644 \u0632\u0645\u0627\u0646 \u0628\u06cc\u0634\u06cc\u0646\u0647 \u06a9\u0646\u062f.<\/p>\n<h3>\u00a0RL \u062f\u0631 \u0645\u062f\u0644\u200c\u0647\u0627\u06cc \u0632\u0628\u0627\u0646\u06cc \u0686\u06af\u0648\u0646\u0647 \u062a\u0639\u0631\u06cc\u0641 \u0645\u06cc\u200c\u0634\u0648\u062f\u061f<\/h3>\n<p>\u0648\u0642\u062a\u06cc \u06cc\u0627\u062f\u06af\u06cc\u0631\u06cc \u062a\u0642\u0648\u06cc\u062a\u06cc \u0631\u0627 \u0628\u0647 \u062f\u0646\u06cc\u0627\u06cc \u0645\u062f\u0644\u200c\u0647\u0627\u06cc \u0632\u0628\u0627\u0646\u06cc \u0628\u0632\u0631\u06af (LLMs) \u0645\u06cc\u200c\u0622\u0648\u0631\u06cc\u0645\u060c \u0627\u06cc\u0646 \u0645\u0641\u0627\u0647\u06cc\u0645 \u0628\u0647 \u0634\u06a9\u0644 \u0632\u06cc\u0631 \u062a\u0641\u0633\u06cc\u0631 \u0645\u06cc\u200c\u0634\u0648\u0646\u062f:<\/p>\n<table style=\"height: 120px\" width=\"1065\">\n<thead>\n<tr>\n<th>\u0645\u0641\u0647\u0648\u0645 \u062f\u0631 RL<\/th>\n<th>\u0645\u0639\u0627\u062f\u0644 \u062f\u0631 \u0645\u062f\u0644 \u0632\u0628\u0627\u0646\u06cc<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td style=\"text-align: center\">Agent<\/td>\n<td style=\"text-align: center\">\u0645\u062f\u0644 \u0632\u0628\u0627\u0646\u06cc<\/td>\n<\/tr>\n<tr>\n<td style=\"text-align: center\">Environment<\/td>\n<td style=\"text-align: center\">\u0645\u06a9\u0627\u0644\u0645\u0647\u060c \u06a9\u0627\u0631\u0628\u0631 \u06cc\u0627 \u06a9\u0627\u0646\u062a\u06a9\u0633\u062a \u0648\u0631\u0648\u062f\u06cc<\/td>\n<\/tr>\n<tr>\n<td style=\"text-align: center\">Action<\/td>\n<td style=\"text-align: center\">\u062a\u0648\u0644\u06cc\u062f \u06cc\u06a9 \u062a\u0648\u06a9\u0646 \u06cc\u0627 \u06cc\u06a9 \u067e\u0627\u0633\u062e \u06a9\u0627\u0645\u0644<\/td>\n<\/tr>\n<tr>\n<td style=\"text-align: center\">Reward<\/td>\n<td style=\"text-align: center\">\u0627\u0645\u062a\u06cc\u0627\u0632 \u06a9\u06cc\u0641\u06cc\u062a\u060c \u0645\u0641\u06cc\u062f \u0628\u0648\u062f\u0646 \u06cc\u0627 \u0645\u0637\u0644\u0648\u0628 \u0628\u0648\u062f\u0646 \u067e\u0627\u0633\u062e<\/td>\n<\/tr>\n<tr>\n<td style=\"text-align: center\">Policy<\/td>\n<td style=\"text-align: center\">\u0627\u062d\u062a\u0645\u0627\u0644 \u062a\u0648\u0644\u06cc\u062f \u062a\u0648\u06a9\u0646\u200c\u0647\u0627 \u062a\u0648\u0633\u0637 \u0645\u062f\u0644<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>\u0628\u0647 \u0627\u06cc\u0646 \u062a\u0631\u062a\u06cc\u0628\u060c \u062a\u0648\u0644\u06cc\u062f \u0645\u062a\u0646 \u062f\u06cc\u06af\u0631 \u0641\u0642\u0637 \u06cc\u06a9 \u0641\u0631\u0622\u06cc\u0646\u062f \u0622\u0645\u0627\u0631\u06cc \u0646\u06cc\u0633\u062a\u060c \u0628\u0644\u06a9\u0647 \u0628\u0647 \u06cc\u06a9 <strong>\u0645\u0633\u0626\u0644\u0647 \u062a\u0635\u0645\u06cc\u0645\u200c\u06af\u06cc\u0631\u06cc<\/strong> \u062a\u0628\u062f\u06cc\u0644 \u0645\u06cc\u200c\u0634\u0648\u062f.<\/p>\n<h3>\u00a0\u0645\u062b\u0627\u0644 \u0633\u0627\u062f\u0647: \u062a\u0641\u0627\u0648\u062a \u0622\u0645\u0648\u0632\u0634 \u0645\u0639\u0645\u0648\u0644\u06cc \u0648 RL<\/h3>\n<p>\u0641\u0631\u0636 \u06a9\u0646\u06cc\u062f \u0627\u0632 \u06cc\u06a9 \u0645\u062f\u0644 \u0632\u0628\u0627\u0646\u06cc \u0627\u06cc\u0646 \u0633\u0624\u0627\u0644 \u0631\u0627 \u0645\u06cc\u200c\u067e\u0631\u0633\u06cc\u0645:<\/p>\n<blockquote><p>\u00ab\u0686\u0637\u0648\u0631 \u0645\u06cc\u200c\u062a\u0648\u0627\u0646\u0645 \u062a\u0645\u0631\u06a9\u0632 \u062e\u0648\u062f\u0645 \u0631\u0627 \u0647\u0646\u06af\u0627\u0645 \u0645\u0637\u0627\u0644\u0639\u0647 \u0627\u0641\u0632\u0627\u06cc\u0634 \u0628\u062f\u0647\u0645\u061f\u00bb<\/p><\/blockquote>\n<h4>\ud83d\udd39 \u0622\u0645\u0648\u0632\u0634 \u0645\u0639\u0645\u0648\u0644\u06cc (Pretraining)<\/h4>\n<p>\u0645\u062f\u0644 \u067e\u0627\u0633\u062e\u06cc \u0645\u06cc\u200c\u062f\u0647\u062f \u06a9\u0647 \u0627\u0632 \u0646\u0638\u0631 \u0622\u0645\u0627\u0631\u06cc \u0634\u0628\u06cc\u0647 \u067e\u0627\u0633\u062e\u200c\u0647\u0627\u06cc \u062f\u06cc\u062f\u0647\u200c\u0634\u062f\u0647 \u062f\u0631 \u062f\u0627\u062f\u0647\u200c\u0647\u0627\u06cc \u0622\u0645\u0648\u0632\u0634\u06cc \u0627\u0633\u062a\u060c \u0627\u0645\u0627:<\/p>\n<ul>\n<li>\u0645\u0645\u06a9\u0646 \u0627\u0633\u062a \u062e\u06cc\u0644\u06cc \u06a9\u0644\u06cc \u0628\u0627\u0634\u062f<\/li>\n<li>\u06cc\u0627 \u0628\u06cc\u0634 \u0627\u0632 \u062d\u062f \u0637\u0648\u0644\u0627\u0646\u06cc<\/li>\n<li>\u06cc\u0627 \u062d\u062a\u06cc \u062a\u0648\u0635\u06cc\u0647\u200c\u0647\u0627\u06cc \u063a\u06cc\u0631\u0639\u0645\u0644\u06cc \u0628\u062f\u0647\u062f<\/li>\n<\/ul>\n<h4>\ud83d\udd39 \u0628\u0627 \u06cc\u0627\u062f\u06af\u06cc\u0631\u06cc \u062a\u0642\u0648\u06cc\u062a\u06cc<\/h4>\n<p>\u0645\u062f\u0644 \u0628\u0631\u0627\u0633\u0627\u0633 \u067e\u0627\u062f\u0627\u0634 \u06cc\u0627\u062f \u0645\u06cc\u200c\u06af\u06cc\u0631\u062f \u06a9\u0647:<\/p>\n<ul>\n<li>\u067e\u0627\u0633\u062e <strong>\u0645\u062e\u062a\u0635\u0631 \u0627\u0645\u0627 \u06a9\u0627\u0631\u0628\u0631\u062f\u06cc<\/strong> \u0628\u062f\u0647\u062f<\/li>\n<li>\u0644\u062d\u0646 \u0645\u0646\u0627\u0633\u0628 \u062f\u0627\u0634\u062a\u0647 \u0628\u0627\u0634\u062f<\/li>\n<li>\u062a\u0648\u0635\u06cc\u0647\u200c\u0647\u0627\u06cc \u0648\u0627\u0642\u0639\u200c\u0628\u06cc\u0646\u0627\u0646\u0647 \u0627\u0631\u0627\u0626\u0647 \u06a9\u0646\u062f<\/li>\n<\/ul>\n<p>\u062f\u0631 \u0627\u06cc\u0646 \u062d\u0627\u0644\u062a\u060c \u0645\u062f\u0644 \u0641\u0642\u0637 \u0646\u0645\u06cc\u200c\u067e\u0631\u0633\u062f \u00ab\u06a9\u062f\u0627\u0645 \u067e\u0627\u0633\u062e \u0645\u062d\u062a\u0645\u0644\u200c\u062a\u0631 \u0627\u0633\u062a\u061f\u00bb<br \/>\n\u0628\u0644\u06a9\u0647 \u0645\u06cc\u200c\u067e\u0631\u0633\u062f:<br \/>\n<strong>\u00ab\u06a9\u062f\u0627\u0645 \u067e\u0627\u0633\u062e \u0628\u0647\u062a\u0631 \u0627\u0633\u062a\u061f\u00bb<\/strong><\/p>\n<hr \/>\n<h3>\u062a\u0641\u0627\u0648\u062a \u06a9\u0644\u06cc\u062f\u06cc RL \u0628\u0627 \u0633\u0627\u06cc\u0631 \u0631\u0648\u0634\u200c\u0647\u0627\u06cc \u06cc\u0627\u062f\u06af\u06cc\u0631\u06cc<\/h3>\n<table style=\"height: 86px\" width=\"1133\">\n<thead>\n<tr>\n<th>\u0631\u0648\u0634 \u06cc\u0627\u062f\u06af\u06cc\u0631\u06cc<\/th>\n<th>\u0645\u0639\u06cc\u0627\u0631 \u0628\u0647\u06cc\u0646\u0647\u200c\u0633\u0627\u0632\u06cc<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td style=\"text-align: center\">\u06cc\u0627\u062f\u06af\u06cc\u0631\u06cc \u0646\u0638\u0627\u0631\u062a\u200c\u0634\u062f\u0647<\/td>\n<td style=\"text-align: center\">\u0634\u0628\u0627\u0647\u062a \u0628\u0647 \u0628\u0631\u0686\u0633\u0628\u200c\u0647\u0627\u06cc \u062f\u0627\u062f\u0647<\/td>\n<\/tr>\n<tr>\n<td style=\"text-align: center\">\u06cc\u0627\u062f\u06af\u06cc\u0631\u06cc \u062e\u0648\u062f\u0646\u0638\u0627\u0631\u062a\u06cc<\/td>\n<td style=\"text-align: center\">\u067e\u06cc\u0634\u200c\u0628\u06cc\u0646\u06cc \u062f\u0642\u06cc\u0642 \u062a\u0648\u06a9\u0646 \u0628\u0639\u062f\u06cc<\/td>\n<\/tr>\n<tr>\n<td style=\"text-align: center\">\u06cc\u0627\u062f\u06af\u06cc\u0631\u06cc \u062a\u0642\u0648\u06cc\u062a\u06cc<\/td>\n<td style=\"text-align: center\">\u0628\u06cc\u0634\u06cc\u0646\u0647\u200c\u0633\u0627\u0632\u06cc \u067e\u0627\u062f\u0627\u0634 \u0631\u0641\u062a\u0627\u0631\u06cc<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>\u0627\u06cc\u0646 \u062a\u0641\u0627\u0648\u062a \u0628\u0627\u0639\u062b \u0645\u06cc\u200c\u0634\u0648\u062f RL \u0627\u0628\u0632\u0627\u0631 \u0645\u0646\u0627\u0633\u0628\u06cc \u0628\u0631\u0627\u06cc <strong>\u06a9\u0646\u062a\u0631\u0644 \u0631\u0641\u062a\u0627\u0631 \u0645\u062f\u0644\u200c\u0647\u0627\u06cc \u0632\u0628\u0627\u0646\u06cc<\/strong> \u0628\u0627\u0634\u062f\u060c \u0646\u0647 \u0641\u0642\u0637 \u0627\u0641\u0632\u0627\u06cc\u0634 \u062f\u0642\u062a \u0622\u0645\u0627\u0631\u06cc \u0622\u0646\u200c\u0647\u0627.<\/p>\n<hr \/>\n<h3>\u00a0\u0686\u0631\u0627 RL \u0628\u0631\u0627\u06cc LLM\u0647\u0627 \u062d\u06cc\u0627\u062a\u06cc \u0627\u0633\u062a\u061f<\/h3>\n<p>\u06cc\u0627\u062f\u06af\u06cc\u0631\u06cc \u062a\u0642\u0648\u06cc\u062a\u06cc \u0628\u0647 \u0645\u062f\u0644\u200c\u0647\u0627\u06cc \u0632\u0628\u0627\u0646\u06cc \u0627\u062c\u0627\u0632\u0647 \u0645\u06cc\u200c\u062f\u0647\u062f:<\/p>\n<ul>\n<li>\u06a9\u06cc\u0641\u06cc\u062a \u067e\u0627\u0633\u062e \u0631\u0627 \u0627\u0632 \u062f\u06cc\u062f \u0627\u0646\u0633\u0627\u0646 \u062f\u0631\u06a9 \u06a9\u0646\u0646\u062f<\/li>\n<li>\u0631\u0641\u062a\u0627\u0631 \u062e\u0648\u062f \u0631\u0627 \u062f\u0631 \u0637\u0648\u0644 \u0645\u06a9\u0627\u0644\u0645\u0647 \u0627\u0635\u0644\u0627\u062d \u06a9\u0646\u0646\u062f<\/li>\n<li>\u0648 \u0628\u0647 \u0627\u0647\u062f\u0627\u0641\u06cc \u0641\u0631\u0627\u062a\u0631 \u0627\u0632 \u00ab\u062f\u0631\u0633\u062a\u200c\u0628\u0648\u062f\u0646 \u0632\u0628\u0627\u0646\u06cc\u00bb \u0628\u0631\u0633\u0646\u062f<\/li>\n<\/ul>\n<p>\u0628\u0647 \u0647\u0645\u06cc\u0646 \u062f\u0644\u06cc\u0644\u060c RL \u067e\u0627\u06cc\u0647\u200c\u06cc \u0628\u0633\u06cc\u0627\u0631\u06cc \u0627\u0632 \u062a\u06a9\u0646\u06cc\u06a9\u200c\u0647\u0627\u06cc \u067e\u06cc\u0634\u0631\u0641\u062a\u0647 \u0645\u0627\u0646\u0646\u062f <strong>RLHF<\/strong> \u0627\u0633\u062a \u06a9\u0647 \u062f\u0631 \u0645\u062f\u0644\u200c\u0647\u0627\u06cc \u0632\u0628\u0627\u0646\u06cc \u0645\u062f\u0631\u0646 \u0627\u0633\u062a\u0641\u0627\u062f\u0647 \u0645\u06cc\u200c\u0634\u0648\u062f.<\/p>\n<p><img fetchpriority=\"high\" decoding=\"async\" class=\"aligncenter wp-image-1240 size-medium_large\" src=\"https:\/\/houshyar24.ir\/blog\/wp-content\/uploads\/2025\/12\/ChatGPT-Image-May-29-2025-09_56_16-AM-768x528.png\" loading=\"lazy\" alt=\"\" width=\"706\" height=\"485\" srcset=\"https:\/\/houshyar24.ir\/blog\/wp-content\/uploads\/2025\/12\/ChatGPT-Image-May-29-2025-09_56_16-AM-768x528.png 768w, https:\/\/houshyar24.ir\/blog\/wp-content\/uploads\/2025\/12\/ChatGPT-Image-May-29-2025-09_56_16-AM-300x206.png 300w, https:\/\/houshyar24.ir\/blog\/wp-content\/uploads\/2025\/12\/ChatGPT-Image-May-29-2025-09_56_16-AM.png 1020w\" sizes=\"(max-width: 706px) 100vw, 706px\" \/><\/p>\n<hr \/>\n<h2>\u00a0\u06cc\u0627\u062f\u06af\u06cc\u0631\u06cc \u062a\u0642\u0648\u06cc\u062a\u06cc \u0628\u0627 \u0628\u0627\u0632\u062e\u0648\u0631\u062f \u0627\u0646\u0633\u0627\u0646\u06cc (RLHF) \u0686\u06cc\u0633\u062a \u0648 \u0686\u0631\u0627 \u0645\u0647\u0645 \u0627\u0633\u062a\u061f<\/h2>\n<p>\u06cc\u06a9\u06cc \u0627\u0632 \u0645\u0647\u0645\u200c\u062a\u0631\u06cc\u0646 \u0648 \u067e\u0631\u06a9\u0627\u0631\u0628\u0631\u062f\u062a\u0631\u06cc\u0646 \u0634\u06a9\u0644\u200c\u0647\u0627\u06cc \u06cc\u0627\u062f\u06af\u06cc\u0631\u06cc \u062a\u0642\u0648\u06cc\u062a\u06cc \u062f\u0631 \u0645\u062f\u0644\u200c\u0647\u0627\u06cc \u0632\u0628\u0627\u0646\u06cc \u0628\u0632\u0631\u06af\u060c <strong>\u06cc\u0627\u062f\u06af\u06cc\u0631\u06cc \u062a\u0642\u0648\u06cc\u062a\u06cc \u0628\u0627 \u0628\u0627\u0632\u062e\u0648\u0631\u062f \u0627\u0646\u0633\u0627\u0646\u06cc<\/strong> \u06cc\u0627 <strong>Reinforcement Learning from Human Feedback (RLHF)<\/strong> \u0627\u0633\u062a.<br \/>\nRLHF \u067e\u0627\u0633\u062e\u06cc \u0645\u0633\u062a\u0642\u06cc\u0645 \u0628\u0647 \u06cc\u06a9 \u0686\u0627\u0644\u0634 \u0627\u0633\u0627\u0633\u06cc \u062f\u0631 LLM\u0647\u0627\u0633\u062a:<\/p>\n<blockquote><p>\u0686\u06af\u0648\u0646\u0647 \u0645\u06cc\u200c\u062a\u0648\u0627\u0646 \u0645\u062f\u0644\u06cc \u0633\u0627\u062e\u062a \u06a9\u0647 \u0646\u0647\u200c\u062a\u0646\u0647\u0627 \u0627\u0632 \u0646\u0638\u0631 \u0632\u0628\u0627\u0646\u06cc \u062f\u0631\u0633\u062a\u060c \u0628\u0644\u06a9\u0647 \u0627\u0632 \u0646\u0638\u0631 \u0627\u0646\u0633\u0627\u0646\u06cc \u00ab\u0645\u0637\u0644\u0648\u0628\u00bb \u0628\u0627\u0634\u062f\u061f<\/p><\/blockquote>\n<hr \/>\n<h3>\u00a0\u0645\u0633\u0626\u0644\u0647\u200c\u0627\u06cc \u06a9\u0647 RLHF \u062d\u0644 \u0645\u06cc\u200c\u06a9\u0646\u062f<\/h3>\n<p>\u0645\u062f\u0644\u200c\u0647\u0627\u06cc \u0632\u0628\u0627\u0646\u06cc \u062f\u0631 \u0645\u0631\u062d\u0644\u0647 \u067e\u06cc\u0634\u200c\u0622\u0645\u0648\u0632\u0634 (Pretraining):<\/p>\n<ul>\n<li>\u0641\u0642\u0637 \u0628\u0631 \u0627\u0633\u0627\u0633 \u0627\u062d\u062a\u0645\u0627\u0644 \u0622\u0645\u0627\u0631\u06cc \u062a\u0648\u06a9\u0646\u200c\u0647\u0627 \u0622\u0645\u0648\u0632\u0634 \u0645\u06cc\u200c\u0628\u06cc\u0646\u0646\u062f<\/li>\n<li>\u062f\u0631\u06a9 \u0645\u0633\u062a\u0642\u06cc\u0645\u06cc \u0627\u0632 \u0645\u0641\u0627\u0647\u06cc\u0645\u06cc \u0645\u062b\u0644 <em>\u0645\u0641\u06cc\u062f \u0628\u0648\u062f\u0646\u060c \u0627\u06cc\u0645\u0646 \u0628\u0648\u062f\u0646 \u06cc\u0627 \u0645\u0624\u062f\u0628 \u0628\u0648\u062f\u0646<\/em> \u0646\u062f\u0627\u0631\u0646\u062f<\/li>\n<li>\u0645\u0645\u06a9\u0646 \u0627\u0633\u062a \u067e\u0627\u0633\u062e\u200c\u0647\u0627\u06cc\u06cc \u062a\u0648\u0644\u06cc\u062f \u06a9\u0646\u0646\u062f \u06a9\u0647 \u0627\u0632 \u0646\u0638\u0631 \u0632\u0628\u0627\u0646\u06cc \u0635\u062d\u06cc\u062d\u060c \u0627\u0645\u0627 \u0627\u0632 \u0646\u0638\u0631 \u0627\u0646\u0633\u0627\u0646\u06cc \u0646\u0627\u0645\u0646\u0627\u0633\u0628 \u0628\u0627\u0634\u0646\u062f<\/li>\n<\/ul>\n<p>RLHF \u0627\u06cc\u0646 \u0634\u06a9\u0627\u0641 \u0631\u0627 \u0628\u0627 \u0648\u0627\u0631\u062f \u06a9\u0631\u062f\u0646 <strong>\u0642\u0636\u0627\u0648\u062a \u0627\u0646\u0633\u0627\u0646\u06cc<\/strong> \u0628\u0647 \u0641\u0631\u0622\u06cc\u0646\u062f \u0622\u0645\u0648\u0632\u0634 \u067e\u0631 \u0645\u06cc\u200c\u06a9\u0646\u062f.<\/p>\n<hr \/>\n<h3>\u00a0\u0641\u0631\u0622\u06cc\u0646\u062f RLHF \u0628\u0647\u200c\u0635\u0648\u0631\u062a \u0645\u0631\u062d\u0644\u0647\u200c\u0628\u0647\u200c\u0645\u0631\u062d\u0644\u0647<\/h3>\n<p>RLHF \u0645\u0639\u0645\u0648\u0644\u0627\u064b \u062f\u0631 \u0633\u0647 \u0645\u0631\u062d\u0644\u0647\u200c\u06cc \u0627\u0635\u0644\u06cc \u0627\u0646\u062c\u0627\u0645 \u0645\u06cc\u200c\u0634\u0648\u062f:<\/p>\n<h4>1\ufe0f\u20e3 \u0622\u0645\u0648\u0632\u0634 \u0645\u062f\u0644 \u067e\u0627\u06cc\u0647 (Pretrained Model)<\/h4>\n<p>\u0645\u062f\u0644 \u0632\u0628\u0627\u0646\u06cc \u0627\u0628\u062a\u062f\u0627 \u0628\u0627 \u0631\u0648\u0634\u200c\u0647\u0627\u06cc \u0631\u0627\u06cc\u062c (\u0645\u062b\u0644 \u067e\u06cc\u0634\u200c\u0628\u06cc\u0646\u06cc \u062a\u0648\u06a9\u0646 \u0628\u0639\u062f\u06cc) \u0631\u0648\u06cc \u062f\u0627\u062f\u0647\u200c\u0647\u0627\u06cc \u0645\u062a\u0646\u06cc \u0628\u0632\u0631\u06af \u0622\u0645\u0648\u0632\u0634 \u062f\u0627\u062f\u0647 \u0645\u06cc\u200c\u0634\u0648\u062f.<br \/>\n\u062f\u0631 \u0627\u06cc\u0646 \u0645\u0631\u062d\u0644\u0647\u060c \u0645\u062f\u0644 \u0632\u0628\u0627\u0646 \u0631\u0627 \u062e\u0648\u0628 \u06cc\u0627\u062f \u0645\u06cc\u200c\u06af\u06cc\u0631\u062f \u0627\u0645\u0627 \u0647\u0646\u0648\u0632 \u0647\u0645\u200c\u0631\u0627\u0633\u062a\u0627\u06cc\u06cc \u0627\u0646\u0633\u0627\u0646\u06cc \u0646\u062f\u0627\u0631\u062f.<\/p>\n<h4>2\ufe0f\u20e3 \u062c\u0645\u0639\u200c\u0622\u0648\u0631\u06cc \u062a\u0631\u062c\u06cc\u062d\u0627\u062a \u0627\u0646\u0633\u0627\u0646\u06cc \u0648 \u0622\u0645\u0648\u0632\u0634 \u0645\u062f\u0644 \u067e\u0627\u062f\u0627\u0634<\/h4>\n<p>\u062f\u0631 \u0627\u06cc\u0646 \u0645\u0631\u062d\u0644\u0647:<\/p>\n<ul>\n<li>\u0645\u062f\u0644 \u0686\u0646\u062f \u067e\u0627\u0633\u062e \u0645\u062e\u062a\u0644\u0641 \u0628\u0647 \u06cc\u06a9 \u067e\u0631\u0633\u0634 \u062a\u0648\u0644\u06cc\u062f \u0645\u06cc\u200c\u06a9\u0646\u062f<\/li>\n<li>\u0627\u0646\u0633\u0627\u0646\u200c\u0647\u0627 \u0627\u06cc\u0646 \u067e\u0627\u0633\u062e\u200c\u0647\u0627 \u0631\u0627 <strong>\u0645\u0642\u0627\u06cc\u0633\u0647 \u06cc\u0627 \u0631\u062a\u0628\u0647\u200c\u0628\u0646\u062f\u06cc<\/strong> \u0645\u06cc\u200c\u06a9\u0646\u0646\u062f<\/li>\n<li>\u062f\u0627\u062f\u0647\u200c\u0647\u0627\u06cc \u062d\u0627\u0635\u0644 \u0628\u0631\u0627\u06cc \u0622\u0645\u0648\u0632\u0634 \u06cc\u06a9 <strong>\u0645\u062f\u0644 \u067e\u0627\u062f\u0627\u0634 (Reward Model)<\/strong> \u0627\u0633\u062a\u0641\u0627\u062f\u0647 \u0645\u06cc\u200c\u0634\u0648\u062f<\/li>\n<\/ul>\n<p>\u0645\u062f\u0644 \u067e\u0627\u062f\u0627\u0634 \u06cc\u0627\u062f \u0645\u06cc\u200c\u06af\u06cc\u0631\u062f \u06a9\u0647:<\/p>\n<blockquote><p>\u06a9\u062f\u0627\u0645 \u067e\u0627\u0633\u062e \u0627\u0632 \u0646\u0638\u0631 \u0627\u0646\u0633\u0627\u0646 \u0628\u0647\u062a\u0631 \u0627\u0633\u062a \u0648 \u0686\u0631\u0627<\/p><\/blockquote>\n<h4>3\ufe0f\u20e3 \u0628\u0647\u06cc\u0646\u0647\u200c\u0633\u0627\u0632\u06cc \u0645\u062f\u0644 \u0628\u0627 \u06cc\u0627\u062f\u06af\u06cc\u0631\u06cc \u062a\u0642\u0648\u06cc\u062a\u06cc<\/h4>\n<p>\u062f\u0631 \u0646\u0647\u0627\u06cc\u062a:<\/p>\n<ul>\n<li>\u0645\u062f\u0644 \u0632\u0628\u0627\u0646\u06cc \u0628\u0647\u200c\u0639\u0646\u0648\u0627\u0646 Agent \u062f\u0631 \u0646\u0638\u0631 \u06af\u0631\u0641\u062a\u0647 \u0645\u06cc\u200c\u0634\u0648\u062f<\/li>\n<li>\u0645\u062f\u0644 \u067e\u0627\u062f\u0627\u0634 \u0646\u0642\u0634 \u0645\u062d\u06cc\u0637 \u0648 \u0645\u0646\u0628\u0639 \u067e\u0627\u062f\u0627\u0634 \u0631\u0627 \u062f\u0627\u0631\u062f<\/li>\n<li>\u0628\u0627 \u0627\u0644\u06af\u0648\u0631\u06cc\u062a\u0645\u200c\u0647\u0627\u06cc\u06cc \u0645\u062b\u0644 <strong>PPO (Proximal Policy Optimization)<\/strong>\u060c \u0645\u062f\u0644 \u0637\u0648\u0631\u06cc \u0628\u0647\u200c\u0631\u0648\u0632\u0631\u0633\u0627\u0646\u06cc \u0645\u06cc\u200c\u0634\u0648\u062f \u06a9\u0647 \u067e\u0627\u062f\u0627\u0634 \u0628\u06cc\u0634\u062a\u0631\u06cc \u0628\u06af\u06cc\u0631\u062f<\/li>\n<\/ul>\n<p>\u0646\u062a\u06cc\u062c\u0647 \u0627\u06cc\u0646 \u0641\u0631\u0622\u06cc\u0646\u062f\u060c \u0645\u062f\u0644\u06cc \u0627\u0633\u062a \u06a9\u0647 \u062e\u0631\u0648\u062c\u06cc\u200c\u0647\u0627\u06cc\u0634 \u0628\u0627 \u062a\u0631\u062c\u06cc\u062d\u0627\u062a \u0627\u0646\u0633\u0627\u0646\u06cc \u0647\u0645\u200c\u0633\u0648\u062a\u0631 \u0627\u0633\u062a.<\/p>\n<hr \/>\n<h3>\u00a0\u0627\u062c\u0632\u0627\u06cc RLHF \u062f\u0631 \u06cc\u06a9 \u0646\u06af\u0627\u0647<\/h3>\n<table style=\"height: 151px\" width=\"1036\">\n<thead>\n<tr>\n<th style=\"text-align: center\">\u0645\u0624\u0644\u0641\u0647<\/th>\n<th style=\"text-align: center\">\u0646\u0642\u0634 \u062f\u0631 RLHF<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td style=\"text-align: center\">\u0645\u062f\u0644 \u0632\u0628\u0627\u0646\u06cc<\/td>\n<td style=\"text-align: center\">\u0639\u0627\u0645\u0644 (Agent)<\/td>\n<\/tr>\n<tr>\n<td style=\"text-align: center\">\u0627\u0646\u0633\u0627\u0646<\/td>\n<td style=\"text-align: center\">\u0645\u0646\u0628\u0639 \u0642\u0636\u0627\u0648\u062a \u0648 \u062a\u0631\u062c\u06cc\u062d<\/td>\n<\/tr>\n<tr>\n<td style=\"text-align: center\">\u0645\u062f\u0644 \u067e\u0627\u062f\u0627\u0634<\/td>\n<td style=\"text-align: center\">\u062a\u0642\u0631\u06cc\u0628 \u0642\u0636\u0627\u0648\u062a \u0627\u0646\u0633\u0627\u0646\u06cc<\/td>\n<\/tr>\n<tr>\n<td style=\"text-align: center\">\u0627\u0644\u06af\u0648\u0631\u06cc\u062a\u0645 RL<\/td>\n<td style=\"text-align: center\">\u0628\u0647\u06cc\u0646\u0647\u200c\u0633\u0627\u0632\u06cc \u0631\u0641\u062a\u0627\u0631 \u0645\u062f\u0644<\/td>\n<\/tr>\n<tr>\n<td style=\"text-align: center\">\u067e\u0627\u062f\u0627\u0634<\/td>\n<td style=\"text-align: center\">\u0645\u06cc\u0632\u0627\u0646 \u0645\u0637\u0644\u0648\u0628 \u0628\u0648\u062f\u0646 \u067e\u0627\u0633\u062e<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>\u0627\u06cc\u0646 \u0645\u0639\u0645\u0627\u0631\u06cc \u0628\u0627\u0639\u062b \u0645\u06cc\u200c\u0634\u0648\u062f RLHF \u0647\u0645 <strong>\u0645\u0642\u06cc\u0627\u0633\u200c\u067e\u0630\u06cc\u0631<\/strong> \u0628\u0627\u0634\u062f \u0648 \u0647\u0645 \u0627\u0632 \u0642\u0636\u0627\u0648\u062a \u0627\u0646\u0633\u0627\u0646\u06cc \u0628\u0647\u0631\u0647 \u0628\u0628\u0631\u062f.<\/p>\n<hr \/>\n<h3>\u00a0\u0645\u062b\u0627\u0644 \u0648\u0627\u0642\u0639\u06cc: \u0686\u0631\u0627 RLHF \u06a9\u06cc\u0641\u06cc\u062a \u067e\u0627\u0633\u062e \u0631\u0627 \u0628\u0647\u062a\u0631 \u0645\u06cc\u200c\u06a9\u0646\u062f\u061f<\/h3>\n<p>\u0641\u0631\u0636 \u06a9\u0646\u06cc\u062f \u067e\u0631\u0633\u0634 \u0632\u06cc\u0631 \u0645\u0637\u0631\u062d \u0634\u0648\u062f:<\/p>\n<blockquote><p>\u00ab\u06cc\u06a9 \u062a\u0648\u0635\u06cc\u0647 \u067e\u0632\u0634\u06a9\u06cc \u0633\u0627\u062f\u0647 \u0628\u062f\u0647.\u00bb<\/p><\/blockquote>\n<h4>\ud83d\udd39 \u0628\u062f\u0648\u0646 RLHF<\/h4>\n<p>\u0645\u062f\u0644 \u0645\u0645\u06a9\u0646 \u0627\u0633\u062a:<\/p>\n<ul>\n<li>\u062a\u0648\u0635\u06cc\u0647\u200c\u0627\u06cc \u06a9\u0644\u06cc\u060c \u0645\u0628\u0647\u0645 \u06cc\u0627 \u062d\u062a\u06cc \u0646\u0627\u062f\u0631\u0633\u062a \u0628\u062f\u0647\u062f<\/li>\n<li>\u0647\u0634\u062f\u0627\u0631\u0647\u0627\u06cc \u0627\u06cc\u0645\u0646\u06cc \u0631\u0627 \u0646\u0627\u062f\u06cc\u062f\u0647 \u0628\u06af\u06cc\u0631\u062f<\/li>\n<li>\u0644\u062d\u0646 \u0646\u0627\u0645\u0646\u0627\u0633\u0628 \u062f\u0627\u0634\u062a\u0647 \u0628\u0627\u0634\u062f<\/li>\n<\/ul>\n<h4>\ud83d\udd39 \u0628\u0627 RLHF<\/h4>\n<p>\u0645\u062f\u0644 \u06cc\u0627\u062f \u0645\u06cc\u200c\u06af\u06cc\u0631\u062f \u06a9\u0647:<\/p>\n<ul>\n<li>\u0627\u0632 \u062a\u0648\u0635\u06cc\u0647\u200c\u0647\u0627\u06cc \u062e\u0637\u0631\u0646\u0627\u06a9 \u067e\u0631\u0647\u06cc\u0632 \u06a9\u0646\u062f<\/li>\n<li>\u0645\u062d\u062f\u0648\u062f\u06cc\u062a\u200c\u0647\u0627\u06cc \u062f\u0627\u0646\u0634 \u062e\u0648\u062f \u0631\u0627 \u0628\u06cc\u0627\u0646 \u06a9\u0646\u062f<\/li>\n<li>\u06a9\u0627\u0631\u0628\u0631 \u0631\u0627 \u0628\u0647 \u0645\u0646\u0627\u0628\u0639 \u0645\u0639\u062a\u0628\u0631 \u06cc\u0627 \u067e\u0632\u0634\u06a9 \u0627\u0631\u062c\u0627\u0639 \u062f\u0647\u062f<\/li>\n<\/ul>\n<p>\u0627\u06cc\u0646 \u062a\u0641\u0627\u0648\u062a\u060c \u0646\u062a\u06cc\u062c\u0647\u200c\u06cc <strong>\u067e\u0627\u062f\u0627\u0634\u200c\u062f\u0647\u06cc \u0628\u0647 \u0631\u0641\u062a\u0627\u0631\u0647\u0627\u06cc \u0627\u06cc\u0645\u0646 \u0648 \u0645\u0633\u0626\u0648\u0644\u0627\u0646\u0647<\/strong> \u0627\u0633\u062a.<\/p>\n<p><img decoding=\"async\" class=\"aligncenter wp-image-1237 size-medium_large\" src=\"https:\/\/houshyar24.ir\/blog\/wp-content\/uploads\/2025\/12\/RLHF-1024x683-1-768x512.webp\" loading=\"lazy\" alt=\"\" width=\"706\" height=\"471\" srcset=\"https:\/\/houshyar24.ir\/blog\/wp-content\/uploads\/2025\/12\/RLHF-1024x683-1-768x512.webp 768w, https:\/\/houshyar24.ir\/blog\/wp-content\/uploads\/2025\/12\/RLHF-1024x683-1-300x200.webp 300w, https:\/\/houshyar24.ir\/blog\/wp-content\/uploads\/2025\/12\/RLHF-1024x683-1.webp 1024w\" sizes=\"(max-width: 706px) 100vw, 706px\" \/><\/p>\n<hr \/>\n<h3>\u00a0\u0686\u0631\u0627 RLHF \u0628\u0631\u0627\u06cc \u0645\u062f\u0644\u200c\u0647\u0627\u06cc \u0632\u0628\u0627\u0646\u06cc \u062d\u06cc\u0627\u062a\u06cc \u0627\u0633\u062a\u061f<\/h3>\n<p>RLHF \u0628\u0647 \u0645\u062f\u0644\u200c\u0647\u0627\u06cc \u0632\u0628\u0627\u0646\u06cc \u06a9\u0645\u06a9 \u0645\u06cc\u200c\u06a9\u0646\u062f \u062a\u0627:<\/p>\n<ul>\n<li>\u0627\u0632 \u062f\u06cc\u062f \u0627\u0646\u0633\u0627\u0646 \u00ab\u067e\u0627\u0633\u062e \u062e\u0648\u0628\u00bb \u0631\u0627 \u06cc\u0627\u062f \u0628\u06af\u06cc\u0631\u0646\u062f<\/li>\n<li>\u0631\u0641\u062a\u0627\u0631 \u062e\u0648\u062f \u0631\u0627 \u0628\u0627 \u06a9\u0627\u0631\u0628\u0631\u062f\u0647\u0627\u06cc \u0648\u0627\u0642\u0639\u06cc \u0633\u0627\u0632\u06af\u0627\u0631 \u06a9\u0646\u0646\u062f<\/li>\n<li>\u0627\u0639\u062a\u0645\u0627\u062f\u067e\u0630\u06cc\u0631\u062a\u0631 \u0648 \u0642\u0627\u0628\u0644 \u0627\u0633\u062a\u0641\u0627\u062f\u0647\u200c\u062a\u0631 \u0634\u0648\u0646\u062f<\/li>\n<\/ul>\n<p>\u0628\u0647 \u0647\u0645\u06cc\u0646 \u062f\u0644\u06cc\u0644\u060c \u0628\u0633\u06cc\u0627\u0631\u06cc \u0627\u0632 \u0645\u062f\u0644\u200c\u0647\u0627\u06cc \u0632\u0628\u0627\u0646\u06cc \u0645\u062f\u0631\u0646 \u062f\u0631 \u0645\u062d\u0635\u0648\u0644\u0627\u062a \u0648\u0627\u0642\u0639\u06cc\u060c \u0627\u0632 RLHF \u0628\u0647\u200c\u0639\u0646\u0648\u0627\u0646 \u0645\u0631\u062d\u0644\u0647\u200c\u06cc \u06a9\u0644\u06cc\u062f\u06cc \u0622\u0645\u0648\u0632\u0634 \u0627\u0633\u062a\u0641\u0627\u062f\u0647 \u0645\u06cc\u200c\u06a9\u0646\u0646\u062f.<\/p>\n<hr \/>\n<h2>\u0631\u0648\u0634\u200c\u0647\u0627\u06cc \u0645\u0631\u062a\u0628\u0637 \u0628\u0627 RLHF: \u0627\u0632 RLAIF \u062a\u0627 DPO<\/h2>\n<p>\u0627\u06af\u0631\u0686\u0647 RLHF \u06cc\u06a9\u06cc \u0627\u0632 \u0631\u0627\u06cc\u062c\u200c\u062a\u0631\u06cc\u0646 \u0631\u0648\u0634\u200c\u0647\u0627\u06cc \u0647\u0645\u200c\u0633\u0648\u0633\u0627\u0632\u06cc \u0645\u062f\u0644\u200c\u0647\u0627\u06cc \u0632\u0628\u0627\u0646\u06cc \u0627\u0633\u062a\u060c \u0627\u0645\u0627 \u0645\u062d\u062f\u0648\u062f\u06cc\u062a\u200c\u0647\u0627\u06cc\u06cc \u0645\u0627\u0646\u0646\u062f \u0647\u0632\u06cc\u0646\u0647 \u0648 \u0645\u0642\u06cc\u0627\u0633\u200c\u067e\u0630\u06cc\u0631\u06cc \u0628\u0627\u0639\u062b \u0634\u062f\u0647 \u0631\u0648\u0634\u200c\u0647\u0627\u06cc \u062c\u0627\u06cc\u06af\u0632\u06cc\u0646 \u0648 \u0645\u06a9\u0645\u0644\u06cc \u062a\u0648\u0633\u0639\u0647 \u067e\u06cc\u062f\u0627 \u06a9\u0646\u0646\u062f.<\/p>\n<h3>\ud83d\udd01 RLAIF: \u06cc\u0627\u062f\u06af\u06cc\u0631\u06cc \u062a\u0642\u0648\u06cc\u062a\u06cc \u0628\u0627 \u0628\u0627\u0632\u062e\u0648\u0631\u062f \u0647\u0648\u0634 \u0645\u0635\u0646\u0648\u0639\u06cc<\/h3>\n<p>\u062f\u0631 <strong>Reinforcement Learning from AI Feedback (RLAIF)<\/strong>\u060c \u0628\u0647\u200c\u062c\u0627\u06cc \u0627\u0646\u0633\u0627\u0646:<\/p>\n<ul>\n<li>\u06cc\u06a9 \u0645\u062f\u0644 \u0632\u0628\u0627\u0646\u06cc \u062f\u06cc\u06af\u0631 \u0646\u0642\u0634 \u0627\u0631\u0632\u06cc\u0627\u0628 \u0631\u0627 \u0628\u0627\u0632\u06cc \u0645\u06cc\u200c\u06a9\u0646\u062f<\/li>\n<li>\u067e\u0627\u0633\u062e\u200c\u0647\u0627 \u0631\u0627 \u0645\u0642\u0627\u06cc\u0633\u0647 \u0648 \u0627\u0645\u062a\u06cc\u0627\u0632\u062f\u0647\u06cc \u0645\u06cc\u200c\u06a9\u0646\u062f<\/li>\n<\/ul>\n<p><strong>\u0645\u0632\u06cc\u062a \u0627\u0635\u0644\u06cc:<\/strong> \u06a9\u0627\u0647\u0634 \u0647\u0632\u06cc\u0646\u0647 \u0648 \u0627\u0641\u0632\u0627\u06cc\u0634 \u0645\u0642\u06cc\u0627\u0633<br \/>\n<strong>\u0686\u0627\u0644\u0634:<\/strong> \u0627\u062d\u062a\u0645\u0627\u0644 \u0627\u0646\u062a\u0642\u0627\u0644 \u0633\u0648\u06af\u06cc\u0631\u06cc\u200c\u0647\u0627\u06cc \u0645\u062f\u0644 \u0627\u0631\u0632\u06cc\u0627\u0628<\/p>\n<h3>\u26a1 DPO: \u0628\u0647\u06cc\u0646\u0647\u200c\u0633\u0627\u0632\u06cc \u0645\u0633\u062a\u0642\u06cc\u0645 \u062a\u0631\u062c\u06cc\u062d\u0627\u062a<\/h3>\n<p><strong>Direct Preference Optimization (DPO)<\/strong> \u0631\u0648\u06cc\u06a9\u0631\u062f\u06cc \u0633\u0627\u062f\u0647\u200c\u062a\u0631 \u0627\u0633\u062a \u06a9\u0647:<\/p>\n<ul>\n<li>\u0645\u0631\u062d\u0644\u0647 \u06a9\u0644\u0627\u0633\u06cc\u06a9 RL \u0631\u0627 \u062d\u0630\u0641 \u0645\u06cc\u200c\u06a9\u0646\u062f<\/li>\n<li>\u0645\u0633\u062a\u0642\u06cc\u0645\u0627\u064b \u0627\u0632 \u062f\u0627\u062f\u0647\u200c\u0647\u0627\u06cc \u062a\u0631\u062c\u06cc\u062d\u06cc \u0628\u0631\u0627\u06cc \u0628\u0647\u06cc\u0646\u0647\u200c\u0633\u0627\u0632\u06cc \u0645\u062f\u0644 \u0627\u0633\u062a\u0641\u0627\u062f\u0647 \u0645\u06cc\u200c\u06a9\u0646\u062f<\/li>\n<\/ul>\n<p>DPO \u062f\u0631 \u0628\u0633\u06cc\u0627\u0631\u06cc \u0627\u0632 \u06a9\u0627\u0631\u0628\u0631\u062f\u0647\u0627:<\/p>\n<ul>\n<li>\u067e\u0627\u06cc\u062f\u0627\u0631\u062a\u0631<\/li>\n<li>\u0633\u0627\u062f\u0647\u200c\u062a\u0631<\/li>\n<li>\u0648 \u0633\u0631\u06cc\u0639\u200c\u062a\u0631 \u0627\u0632 RLHF \u0639\u0645\u0644 \u0645\u06cc\u200c\u06a9\u0646\u062f<\/li>\n<\/ul>\n<p><img decoding=\"async\" class=\"aligncenter wp-image-1238 size-full\" src=\"https:\/\/houshyar24.ir\/blog\/wp-content\/uploads\/2025\/12\/G2CM_FI643_Learn_Article_Images_RLHF_Models_V1a.webp\" loading=\"lazy\" alt=\"\" width=\"690\" height=\"400\" srcset=\"https:\/\/houshyar24.ir\/blog\/wp-content\/uploads\/2025\/12\/G2CM_FI643_Learn_Article_Images_RLHF_Models_V1a.webp 690w, https:\/\/houshyar24.ir\/blog\/wp-content\/uploads\/2025\/12\/G2CM_FI643_Learn_Article_Images_RLHF_Models_V1a-300x174.webp 300w\" sizes=\"(max-width: 690px) 100vw, 690px\" \/><\/p>\n<hr \/>\n<h2>\ud83d\udcc8 \u0645\u0632\u0627\u06cc\u0627 \u0648 \u0686\u0627\u0644\u0634\u200c\u0647\u0627\u06cc \u0627\u0633\u062a\u0641\u0627\u062f\u0647 \u0627\u0632 RL \u062f\u0631 \u0645\u062f\u0644\u200c\u0647\u0627\u06cc \u0632\u0628\u0627\u0646\u06cc<\/h2>\n<h3>\u2705 \u0645\u0632\u0627\u06cc\u0627<\/h3>\n<ul>\n<li>\u0647\u0645\u200c\u0633\u0648\u06cc\u06cc \u0628\u0647\u062a\u0631 \u0628\u0627 \u062a\u0631\u062c\u06cc\u062d\u0627\u062a \u0627\u0646\u0633\u0627\u0646\u06cc<\/li>\n<li>\u0628\u0647\u0628\u0648\u062f \u06a9\u06cc\u0641\u06cc\u062a \u067e\u0627\u0633\u062e \u062f\u0631 \u06a9\u0627\u0631\u0628\u0631\u062f\u0647\u0627\u06cc \u0648\u0627\u0642\u0639\u06cc<\/li>\n<li>\u0627\u0641\u0632\u0627\u06cc\u0634 \u0627\u06cc\u0645\u0646\u06cc \u0648 \u06a9\u0646\u062a\u0631\u0644\u200c\u067e\u0630\u06cc\u0631\u06cc \u0645\u062f\u0644<\/li>\n<\/ul>\n<h3>\u26a0\ufe0f \u0686\u0627\u0644\u0634\u200c\u0647\u0627<\/h3>\n<ul>\n<li>\u062a\u0639\u0631\u06cc\u0641 \u062f\u0642\u06cc\u0642 \u00ab\u067e\u0627\u062f\u0627\u0634 \u062e\u0648\u0628\u00bb<\/li>\n<li>\u062d\u0633\u0627\u0633\u06cc\u062a \u0628\u0647 \u062f\u0627\u062f\u0647\u200c\u0647\u0627\u06cc \u062a\u0631\u062c\u06cc\u062d\u06cc<\/li>\n<li>\u062e\u0637\u0631 \u0628\u06cc\u0634\u200c\u0628\u0631\u0627\u0632\u0634 \u0628\u0647 \u0633\u0644\u06cc\u0642\u0647\u200c\u0647\u0627\u06cc \u062e\u0627\u0635<\/li>\n<\/ul>\n<p>\u0628\u0647 \u0647\u0645\u06cc\u0646 \u062f\u0644\u06cc\u0644\u060c \u0627\u0633\u062a\u0641\u0627\u062f\u0647 \u0627\u0632 RL \u0646\u06cc\u0627\u0632\u0645\u0646\u062f \u0637\u0631\u0627\u062d\u06cc \u062f\u0642\u06cc\u0642 \u0648 \u0627\u0631\u0632\u06cc\u0627\u0628\u06cc \u0645\u062f\u0627\u0648\u0645 \u0627\u0633\u062a.<\/p>\n<hr \/>\n<h2>\u00a0\u0645\u062b\u0627\u0644 \u06a9\u0627\u0631\u0628\u0631\u062f\u06cc: RL \u062f\u0631 \u0645\u062d\u0635\u0648\u0644\u0627\u062a \u0648\u0627\u0642\u0639\u06cc<\/h2>\n<p>\u062f\u0631 \u0686\u062a\u200c\u0628\u0627\u062a\u200c\u0647\u0627 \u0648 \u062f\u0633\u062a\u06cc\u0627\u0631\u0647\u0627\u06cc \u0647\u0648\u0634\u0645\u0646\u062f:<\/p>\n<ul>\n<li>RL \u0628\u0631\u0627\u06cc \u0628\u0647\u0628\u0648\u062f \u0644\u062d\u0646\u060c \u0645\u0641\u06cc\u062f \u0628\u0648\u062f\u0646 \u0648 \u0627\u06cc\u0645\u0646\u06cc \u067e\u0627\u0633\u062e\u200c\u0647\u0627 \u0627\u0633\u062a\u0641\u0627\u062f\u0647 \u0645\u06cc\u200c\u0634\u0648\u062f<\/li>\n<li>\u0645\u062f\u0644 \u06cc\u0627\u062f \u0645\u06cc\u200c\u06af\u06cc\u0631\u062f \u0686\u0647 \u0632\u0645\u0627\u0646\u06cc \u067e\u0627\u0633\u062e \u0646\u062f\u0647\u062f \u06cc\u0627 \u0647\u0634\u062f\u0627\u0631 \u0628\u062f\u0647\u062f<\/li>\n<li>\u06a9\u06cc\u0641\u06cc\u062a \u062a\u062c\u0631\u0628\u0647 \u06a9\u0627\u0631\u0628\u0631 \u0628\u0647\u200c\u0645\u0631\u0648\u0631 \u0627\u0641\u0632\u0627\u06cc\u0634 \u067e\u06cc\u062f\u0627 \u0645\u06cc\u200c\u06a9\u0646\u062f<\/li>\n<\/ul>\n<p>\u0627\u06cc\u0646 \u0646\u0648\u0639 \u0628\u0647\u06cc\u0646\u0647\u200c\u0633\u0627\u0632\u06cc \u0628\u062f\u0648\u0646 \u06cc\u0627\u062f\u06af\u06cc\u0631\u06cc \u062a\u0642\u0648\u06cc\u062a\u06cc \u0639\u0645\u0644\u0627\u064b \u0627\u0645\u06a9\u0627\u0646\u200c\u067e\u0630\u06cc\u0631 \u0646\u06cc\u0633\u062a.<\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter wp-image-1239\" src=\"https:\/\/houshyar24.ir\/blog\/wp-content\/uploads\/2025\/12\/dfeb866c-ab94-4853-be2b-030ae535bc11_TopToolsforRLHF-Encord.avif\" loading=\"lazy\" alt=\"\" width=\"750\" height=\"481\" \/><\/p>\n<hr \/>\n<h2>\u00a0\u062c\u0645\u0639\u200c\u0628\u0646\u062f\u06cc<\/h2>\n<p>\u06cc\u0627\u062f\u06af\u06cc\u0631\u06cc \u062a\u0642\u0648\u06cc\u062a\u06cc\u060c \u0628\u0647\u200c\u0648\u06cc\u0698\u0647 \u062f\u0631 \u0642\u0627\u0644\u0628 RLHF \u0648 \u0631\u0648\u0634\u200c\u0647\u0627\u06cc \u062c\u062f\u06cc\u062f\u062a\u0631 \u0645\u0627\u0646\u0646\u062f DPO\u060c \u0646\u0642\u0634 \u06a9\u0644\u06cc\u062f\u06cc \u062f\u0631 \u062a\u0628\u062f\u06cc\u0644 \u0645\u062f\u0644\u200c\u0647\u0627\u06cc \u0632\u0628\u0627\u0646\u06cc \u0628\u0647 \u0633\u06cc\u0633\u062a\u0645\u200c\u0647\u0627\u06cc\u06cc <strong>\u0642\u0627\u0628\u0644 \u0627\u0639\u062a\u0645\u0627\u062f\u060c \u0647\u0645\u200c\u0633\u0648 \u0628\u0627 \u0627\u0646\u0633\u0627\u0646 \u0648 \u06a9\u0627\u0631\u0628\u0631\u062f\u06cc<\/strong> \u062f\u0627\u0631\u062f.<\/p>\n<p>\u062f\u0631 \u0622\u06cc\u0646\u062f\u0647\u060c \u0627\u0646\u062a\u0638\u0627\u0631 \u0645\u06cc\u200c\u0631\u0648\u062f:<\/p>\n<ul>\n<li>\u0627\u062a\u06a9\u0627 \u0628\u0647 \u0628\u0627\u0632\u062e\u0648\u0631\u062f \u0627\u0646\u0633\u0627\u0646\u06cc \u06a9\u0645\u062a\u0631 \u0634\u0648\u062f<\/li>\n<li>\u0631\u0648\u0634\u200c\u0647\u0627\u06cc \u0633\u0627\u062f\u0647\u200c\u062a\u0631 \u0648 \u0645\u0642\u06cc\u0627\u0633\u200c\u067e\u0630\u06cc\u0631\u062a\u0631 \u0631\u0648\u0627\u062c \u067e\u06cc\u062f\u0627 \u06a9\u0646\u0646\u062f<\/li>\n<li>RL \u0628\u0647 \u0628\u062e\u0634\u06cc \u062c\u062f\u0627\u0646\u0634\u062f\u0646\u06cc \u0627\u0632 \u0622\u0645\u0648\u0632\u0634 LLM\u0647\u0627 \u062a\u0628\u062f\u06cc\u0644 \u0634\u0648\u062f<\/li>\n<\/ul>\n<p>\u062f\u0631 \u0646\u0647\u0627\u06cc\u062a\u060c RL \u067e\u0644\u06cc \u0627\u0633\u062a \u0628\u06cc\u0646 \u00ab\u0645\u062f\u0644\u06cc \u06a9\u0647 \u0632\u0628\u0627\u0646 \u0631\u0627 \u0628\u0644\u062f \u0627\u0633\u062a\u00bb \u0648 \u00ab\u0645\u062f\u0644\u06cc \u06a9\u0647 \u0631\u0641\u062a\u0627\u0631 \u062f\u0631\u0633\u062a \u062f\u0627\u0631\u062f\u00bb.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>\u0645\u062f\u0644\u200c\u0647\u0627\u06cc \u0632\u0628\u0627\u0646\u06cc \u0628\u0632\u0631\u06af (Large Language Models \u06cc\u0627 LLMs) \u062f\u0631 \u0633\u0627\u0644\u200c\u0647\u0627\u06cc \u0627\u062e\u06cc\u0631 \u067e\u06cc\u0634\u0631\u0641\u062a \u0686\u0634\u0645\u06af\u06cc\u0631\u06cc \u062f\u0627\u0634\u062a\u0647\u200c\u0627\u0646\u062f \u0648 \u0628\u0647 \u0627\u0628\u0632\u0627\u0631\u0647\u0627\u06cc\u06cc \u0642\u062f\u0631\u062a\u0645\u0646\u062f \u0628\u0631\u0627\u06cc \u062a\u0648\u0644\u06cc\u062f \u0645\u062a\u0646\u060c \u067e\u0627\u0633\u062e \u0628\u0647 \u067e\u0631\u0633\u0634\u200c\u0647\u0627 \u0648&#8230;<\/p>\n","protected":false},"author":5,"featured_media":1241,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_themeisle_gutenberg_block_has_review":false,"footnotes":""},"categories":[5],"tags":[],"class_list":["post-1236","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-5"],"yoast_head":"<!-- This site is optimized with the Yoast SEO Premium plugin v26.0 (Yoast SEO v26.0) - https:\/\/yoast.com\/wordpress\/plugins\/seo\/ -->\n<title>\u00a0\u06cc\u0627\u062f\u06af\u06cc\u0631\u06cc \u062a\u0642\u0648\u06cc\u062a\u06cc \u062f\u0631 \u0645\u062f\u0644\u200c\u0647\u0627\u06cc \u0632\u0628\u0627\u0646\u06cc: \u0686\u0631\u0627 RL \u0642\u0644\u0628 \u062a\u067e\u0646\u062f\u0647 LLM\u0647\u0627\u06cc \u0645\u062f\u0631\u0646 \u0627\u0633\u062a\u061f - \u0648\u0628\u0644\u0627\u06af \u0647\u0648\u0634\u200c\u06cc\u0627\u063124<\/title>\n<meta name=\"description\" content=\"\u06cc\u0627\u062f\u06af\u06cc\u0631\u06cc \u062a\u0642\u0648\u06cc\u062a\u06cc \u062f\u0631 \u0645\u062f\u0644\u200c\u0647\u0627\u06cc \u0632\u0628\u0627\u0646\u06cc \u0686\u06cc\u0633\u062a \u0648 \u0686\u0631\u0627 \u0628\u0631\u0627\u06cc LLM\u0647\u0627\u06cc \u0645\u062f\u0631\u0646 \u062d\u06cc\u0627\u062a\u06cc \u0627\u0633\u062a\u061f \u062f\u0631 \u0627\u06cc\u0646 \u0645\u0642\u0627\u0644\u0647 \u0628\u0627 RL\u060c RLHF \u0648 \u0631\u0648\u0634\u200c\u0647\u0627\u06cc \u062c\u062f\u06cc\u062f \u0647\u0645\u200c\u0633\u0648\u0633\u0627\u0632\u06cc \u0645\u062f\u0644\u200c\u0647\u0627\u06cc \u0632\u0628\u0627\u0646\u06cc \u0628\u0647\u200c\u0635\u0648\u0631\u062a \u0639\u0644\u0645\u06cc \u0648 \u0642\u0627\u0628\u0644 \u0641\u0647\u0645 \u0622\u0634\u0646\u0627 \u0645\u06cc\u200c\u0634\u0648\u06cc\u062f.\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/houshyar24.ir\/blog\/reinforcement-learning-in-llms\/\" \/>\n<meta property=\"og:locale\" content=\"fa_IR\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"\u00a0\u06cc\u0627\u062f\u06af\u06cc\u0631\u06cc \u062a\u0642\u0648\u06cc\u062a\u06cc \u062f\u0631 \u0645\u062f\u0644\u200c\u0647\u0627\u06cc \u0632\u0628\u0627\u0646\u06cc: \u0686\u0631\u0627 RL \u0642\u0644\u0628 \u062a\u067e\u0646\u062f\u0647 LLM\u0647\u0627\u06cc \u0645\u062f\u0631\u0646 \u0627\u0633\u062a\u061f\" \/>\n<meta property=\"og:description\" content=\"\u06cc\u0627\u062f\u06af\u06cc\u0631\u06cc \u062a\u0642\u0648\u06cc\u062a\u06cc \u062f\u0631 \u0645\u062f\u0644\u200c\u0647\u0627\u06cc \u0632\u0628\u0627\u0646\u06cc \u0686\u06cc\u0633\u062a \u0648 \u0686\u0631\u0627 \u0628\u0631\u0627\u06cc LLM\u0647\u0627\u06cc \u0645\u062f\u0631\u0646 \u062d\u06cc\u0627\u062a\u06cc \u0627\u0633\u062a\u061f \u062f\u0631 \u0627\u06cc\u0646 \u0645\u0642\u0627\u0644\u0647 \u0628\u0627 RL\u060c RLHF \u0648 \u0631\u0648\u0634\u200c\u0647\u0627\u06cc \u062c\u062f\u06cc\u062f \u0647\u0645\u200c\u0633\u0648\u0633\u0627\u0632\u06cc \u0645\u062f\u0644\u200c\u0647\u0627\u06cc \u0632\u0628\u0627\u0646\u06cc \u0628\u0647\u200c\u0635\u0648\u0631\u062a \u0639\u0644\u0645\u06cc \u0648 \u0642\u0627\u0628\u0644 \u0641\u0647\u0645 \u0622\u0634\u0646\u0627 \u0645\u06cc\u200c\u0634\u0648\u06cc\u062f.\" \/>\n<meta property=\"og:url\" content=\"https:\/\/houshyar24.ir\/blog\/reinforcement-learning-in-llms\/\" \/>\n<meta property=\"og:site_name\" content=\"\u0648\u0628\u0644\u0627\u06af \u0647\u0648\u0634\u200c\u06cc\u0627\u063124\" \/>\n<meta property=\"article:published_time\" content=\"2025-12-31T16:54:55+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/houshyar24.ir\/blog\/wp-content\/uploads\/2025\/12\/f67c7f1c-0fc5-4afd-a305-0c3086fd909c-1024x683.png\" \/>\n\t<meta property=\"og:image:width\" content=\"1024\" \/>\n\t<meta property=\"og:image:height\" content=\"683\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/png\" \/>\n<meta name=\"author\" content=\"\u0641\u0627\u0626\u0632\u0647 \u0639\u0628\u062f\u0644\u06cc \u0646\u0698\u0627\u062f\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"\u0646\u0648\u0634\u062a\u0647\u200c\u0634\u062f\u0647 \u0628\u062f\u0633\u062a\" \/>\n\t<meta name=\"twitter:data1\" content=\"\u0641\u0627\u0626\u0632\u0647 \u0639\u0628\u062f\u0644\u06cc \u0646\u0698\u0627\u062f\" \/>\n\t<meta name=\"twitter:label2\" content=\"\u0632\u0645\u0627\u0646 \u062a\u0642\u0631\u06cc\u0628\u06cc \u0628\u0631\u0627\u06cc \u062e\u0648\u0627\u0646\u062f\u0646\" \/>\n\t<meta name=\"twitter:data2\" content=\"8 \u062f\u0642\u06cc\u0642\u0647\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"WebPage\",\"@id\":\"https:\/\/houshyar24.ir\/blog\/reinforcement-learning-in-llms\/\",\"url\":\"https:\/\/houshyar24.ir\/blog\/reinforcement-learning-in-llms\/\",\"name\":\"\u00a0\u06cc\u0627\u062f\u06af\u06cc\u0631\u06cc \u062a\u0642\u0648\u06cc\u062a\u06cc \u062f\u0631 \u0645\u062f\u0644\u200c\u0647\u0627\u06cc \u0632\u0628\u0627\u0646\u06cc: \u0686\u0631\u0627 RL \u0642\u0644\u0628 \u062a\u067e\u0646\u062f\u0647 LLM\u0647\u0627\u06cc \u0645\u062f\u0631\u0646 \u0627\u0633\u062a\u061f - \u0648\u0628\u0644\u0627\u06af \u0647\u0648\u0634\u200c\u06cc\u0627\u063124\",\"isPartOf\":{\"@id\":\"https:\/\/houshyar24.ir\/blog\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\/\/houshyar24.ir\/blog\/reinforcement-learning-in-llms\/#primaryimage\"},\"image\":{\"@id\":\"https:\/\/houshyar24.ir\/blog\/reinforcement-learning-in-llms\/#primaryimage\"},\"thumbnailUrl\":\"https:\/\/houshyar24.ir\/blog\/wp-content\/uploads\/2025\/12\/f67c7f1c-0fc5-4afd-a305-0c3086fd909c.png\",\"datePublished\":\"2025-12-31T16:54:55+00:00\",\"author\":{\"@id\":\"https:\/\/houshyar24.ir\/blog\/#\/schema\/person\/5136eb5d0fe45d207dbfbf8f24ae5217\"},\"description\":\"\u06cc\u0627\u062f\u06af\u06cc\u0631\u06cc \u062a\u0642\u0648\u06cc\u062a\u06cc \u062f\u0631 \u0645\u062f\u0644\u200c\u0647\u0627\u06cc \u0632\u0628\u0627\u0646\u06cc \u0686\u06cc\u0633\u062a \u0648 \u0686\u0631\u0627 \u0628\u0631\u0627\u06cc LLM\u0647\u0627\u06cc \u0645\u062f\u0631\u0646 \u062d\u06cc\u0627\u062a\u06cc \u0627\u0633\u062a\u061f \u062f\u0631 \u0627\u06cc\u0646 \u0645\u0642\u0627\u0644\u0647 \u0628\u0627 RL\u060c RLHF \u0648 \u0631\u0648\u0634\u200c\u0647\u0627\u06cc \u062c\u062f\u06cc\u062f \u0647\u0645\u200c\u0633\u0648\u0633\u0627\u0632\u06cc \u0645\u062f\u0644\u200c\u0647\u0627\u06cc \u0632\u0628\u0627\u0646\u06cc \u0628\u0647\u200c\u0635\u0648\u0631\u062a \u0639\u0644\u0645\u06cc \u0648 \u0642\u0627\u0628\u0644 \u0641\u0647\u0645 \u0622\u0634\u0646\u0627 \u0645\u06cc\u200c\u0634\u0648\u06cc\u062f.\",\"breadcrumb\":{\"@id\":\"https:\/\/houshyar24.ir\/blog\/reinforcement-learning-in-llms\/#breadcrumb\"},\"inLanguage\":\"fa-IR\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/houshyar24.ir\/blog\/reinforcement-learning-in-llms\/\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"fa-IR\",\"@id\":\"https:\/\/houshyar24.ir\/blog\/reinforcement-learning-in-llms\/#primaryimage\",\"url\":\"https:\/\/houshyar24.ir\/blog\/wp-content\/uploads\/2025\/12\/f67c7f1c-0fc5-4afd-a305-0c3086fd909c.png\",\"contentUrl\":\"https:\/\/houshyar24.ir\/blog\/wp-content\/uploads\/2025\/12\/f67c7f1c-0fc5-4afd-a305-0c3086fd909c.png\",\"width\":1536,\"height\":1024},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\/\/houshyar24.ir\/blog\/reinforcement-learning-in-llms\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"\u062e\u0627\u0646\u0647\",\"item\":\"https:\/\/houshyar24.ir\/blog\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"\u00a0\u06cc\u0627\u062f\u06af\u06cc\u0631\u06cc \u062a\u0642\u0648\u06cc\u062a\u06cc \u062f\u0631 \u0645\u062f\u0644\u200c\u0647\u0627\u06cc \u0632\u0628\u0627\u0646\u06cc: \u0686\u0631\u0627 RL \u0642\u0644\u0628 \u062a\u067e\u0646\u062f\u0647 LLM\u0647\u0627\u06cc \u0645\u062f\u0631\u0646 \u0627\u0633\u062a\u061f\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/houshyar24.ir\/blog\/#website\",\"url\":\"https:\/\/houshyar24.ir\/blog\/\",\"name\":\"\u0648\u0628\u0644\u0627\u06af \u0647\u0648\u0634\u200c\u06cc\u0627\u063124\",\"description\":\"\",\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/houshyar24.ir\/blog\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"fa-IR\"},{\"@type\":\"Person\",\"@id\":\"https:\/\/houshyar24.ir\/blog\/#\/schema\/person\/5136eb5d0fe45d207dbfbf8f24ae5217\",\"name\":\"\u0641\u0627\u0626\u0632\u0647 \u0639\u0628\u062f\u0644\u06cc \u0646\u0698\u0627\u062f\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"fa-IR\",\"@id\":\"https:\/\/houshyar24.ir\/blog\/#\/schema\/person\/image\/\",\"url\":\"https:\/\/secure.gravatar.com\/avatar\/bbaebbb93cd807693dd89d77e905119940677df168b33f59e6b15c153249fac1?s=96&d=mm&r=g\",\"contentUrl\":\"https:\/\/secure.gravatar.com\/avatar\/bbaebbb93cd807693dd89d77e905119940677df168b33f59e6b15c153249fac1?s=96&d=mm&r=g\",\"caption\":\"\u0641\u0627\u0626\u0632\u0647 \u0639\u0628\u062f\u0644\u06cc \u0646\u0698\u0627\u062f\"},\"url\":\"https:\/\/houshyar24.ir\/blog\/author\/faeze\/\"}]}<\/script>\n<!-- \/ Yoast SEO Premium plugin. -->","yoast_head_json":{"title":"\u00a0\u06cc\u0627\u062f\u06af\u06cc\u0631\u06cc \u062a\u0642\u0648\u06cc\u062a\u06cc \u062f\u0631 \u0645\u062f\u0644\u200c\u0647\u0627\u06cc \u0632\u0628\u0627\u0646\u06cc: \u0686\u0631\u0627 RL \u0642\u0644\u0628 \u062a\u067e\u0646\u062f\u0647 LLM\u0647\u0627\u06cc \u0645\u062f\u0631\u0646 \u0627\u0633\u062a\u061f - \u0648\u0628\u0644\u0627\u06af \u0647\u0648\u0634\u200c\u06cc\u0627\u063124","description":"\u06cc\u0627\u062f\u06af\u06cc\u0631\u06cc \u062a\u0642\u0648\u06cc\u062a\u06cc \u062f\u0631 \u0645\u062f\u0644\u200c\u0647\u0627\u06cc \u0632\u0628\u0627\u0646\u06cc \u0686\u06cc\u0633\u062a \u0648 \u0686\u0631\u0627 \u0628\u0631\u0627\u06cc LLM\u0647\u0627\u06cc \u0645\u062f\u0631\u0646 \u062d\u06cc\u0627\u062a\u06cc \u0627\u0633\u062a\u061f \u062f\u0631 \u0627\u06cc\u0646 \u0645\u0642\u0627\u0644\u0647 \u0628\u0627 RL\u060c RLHF \u0648 \u0631\u0648\u0634\u200c\u0647\u0627\u06cc \u062c\u062f\u06cc\u062f \u0647\u0645\u200c\u0633\u0648\u0633\u0627\u0632\u06cc \u0645\u062f\u0644\u200c\u0647\u0627\u06cc \u0632\u0628\u0627\u0646\u06cc \u0628\u0647\u200c\u0635\u0648\u0631\u062a \u0639\u0644\u0645\u06cc \u0648 \u0642\u0627\u0628\u0644 \u0641\u0647\u0645 \u0622\u0634\u0646\u0627 \u0645\u06cc\u200c\u0634\u0648\u06cc\u062f.","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/houshyar24.ir\/blog\/reinforcement-learning-in-llms\/","og_locale":"fa_IR","og_type":"article","og_title":"\u00a0\u06cc\u0627\u062f\u06af\u06cc\u0631\u06cc \u062a\u0642\u0648\u06cc\u062a\u06cc \u062f\u0631 \u0645\u062f\u0644\u200c\u0647\u0627\u06cc \u0632\u0628\u0627\u0646\u06cc: \u0686\u0631\u0627 RL \u0642\u0644\u0628 \u062a\u067e\u0646\u062f\u0647 LLM\u0647\u0627\u06cc \u0645\u062f\u0631\u0646 \u0627\u0633\u062a\u061f","og_description":"\u06cc\u0627\u062f\u06af\u06cc\u0631\u06cc \u062a\u0642\u0648\u06cc\u062a\u06cc \u062f\u0631 \u0645\u062f\u0644\u200c\u0647\u0627\u06cc \u0632\u0628\u0627\u0646\u06cc \u0686\u06cc\u0633\u062a \u0648 \u0686\u0631\u0627 \u0628\u0631\u0627\u06cc LLM\u0647\u0627\u06cc \u0645\u062f\u0631\u0646 \u062d\u06cc\u0627\u062a\u06cc \u0627\u0633\u062a\u061f \u062f\u0631 \u0627\u06cc\u0646 \u0645\u0642\u0627\u0644\u0647 \u0628\u0627 RL\u060c RLHF \u0648 \u0631\u0648\u0634\u200c\u0647\u0627\u06cc \u062c\u062f\u06cc\u062f \u0647\u0645\u200c\u0633\u0648\u0633\u0627\u0632\u06cc \u0645\u062f\u0644\u200c\u0647\u0627\u06cc \u0632\u0628\u0627\u0646\u06cc \u0628\u0647\u200c\u0635\u0648\u0631\u062a \u0639\u0644\u0645\u06cc \u0648 \u0642\u0627\u0628\u0644 \u0641\u0647\u0645 \u0622\u0634\u0646\u0627 \u0645\u06cc\u200c\u0634\u0648\u06cc\u062f.","og_url":"https:\/\/houshyar24.ir\/blog\/reinforcement-learning-in-llms\/","og_site_name":"\u0648\u0628\u0644\u0627\u06af \u0647\u0648\u0634\u200c\u06cc\u0627\u063124","article_published_time":"2025-12-31T16:54:55+00:00","og_image":[{"width":1024,"height":683,"url":"https:\/\/houshyar24.ir\/blog\/wp-content\/uploads\/2025\/12\/f67c7f1c-0fc5-4afd-a305-0c3086fd909c-1024x683.png","type":"image\/png"}],"author":"\u0641\u0627\u0626\u0632\u0647 \u0639\u0628\u062f\u0644\u06cc \u0646\u0698\u0627\u062f","twitter_card":"summary_large_image","twitter_misc":{"\u0646\u0648\u0634\u062a\u0647\u200c\u0634\u062f\u0647 \u0628\u062f\u0633\u062a":"\u0641\u0627\u0626\u0632\u0647 \u0639\u0628\u062f\u0644\u06cc \u0646\u0698\u0627\u062f","\u0632\u0645\u0627\u0646 \u062a\u0642\u0631\u06cc\u0628\u06cc \u0628\u0631\u0627\u06cc \u062e\u0648\u0627\u0646\u062f\u0646":"8 \u062f\u0642\u06cc\u0642\u0647"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"WebPage","@id":"https:\/\/houshyar24.ir\/blog\/reinforcement-learning-in-llms\/","url":"https:\/\/houshyar24.ir\/blog\/reinforcement-learning-in-llms\/","name":"\u00a0\u06cc\u0627\u062f\u06af\u06cc\u0631\u06cc \u062a\u0642\u0648\u06cc\u062a\u06cc \u062f\u0631 \u0645\u062f\u0644\u200c\u0647\u0627\u06cc \u0632\u0628\u0627\u0646\u06cc: \u0686\u0631\u0627 RL \u0642\u0644\u0628 \u062a\u067e\u0646\u062f\u0647 LLM\u0647\u0627\u06cc \u0645\u062f\u0631\u0646 \u0627\u0633\u062a\u061f - \u0648\u0628\u0644\u0627\u06af \u0647\u0648\u0634\u200c\u06cc\u0627\u063124","isPartOf":{"@id":"https:\/\/houshyar24.ir\/blog\/#website"},"primaryImageOfPage":{"@id":"https:\/\/houshyar24.ir\/blog\/reinforcement-learning-in-llms\/#primaryimage"},"image":{"@id":"https:\/\/houshyar24.ir\/blog\/reinforcement-learning-in-llms\/#primaryimage"},"thumbnailUrl":"https:\/\/houshyar24.ir\/blog\/wp-content\/uploads\/2025\/12\/f67c7f1c-0fc5-4afd-a305-0c3086fd909c.png","datePublished":"2025-12-31T16:54:55+00:00","author":{"@id":"https:\/\/houshyar24.ir\/blog\/#\/schema\/person\/5136eb5d0fe45d207dbfbf8f24ae5217"},"description":"\u06cc\u0627\u062f\u06af\u06cc\u0631\u06cc \u062a\u0642\u0648\u06cc\u062a\u06cc \u062f\u0631 \u0645\u062f\u0644\u200c\u0647\u0627\u06cc \u0632\u0628\u0627\u0646\u06cc \u0686\u06cc\u0633\u062a \u0648 \u0686\u0631\u0627 \u0628\u0631\u0627\u06cc LLM\u0647\u0627\u06cc \u0645\u062f\u0631\u0646 \u062d\u06cc\u0627\u062a\u06cc \u0627\u0633\u062a\u061f \u062f\u0631 \u0627\u06cc\u0646 \u0645\u0642\u0627\u0644\u0647 \u0628\u0627 RL\u060c RLHF \u0648 \u0631\u0648\u0634\u200c\u0647\u0627\u06cc \u062c\u062f\u06cc\u062f \u0647\u0645\u200c\u0633\u0648\u0633\u0627\u0632\u06cc \u0645\u062f\u0644\u200c\u0647\u0627\u06cc \u0632\u0628\u0627\u0646\u06cc \u0628\u0647\u200c\u0635\u0648\u0631\u062a \u0639\u0644\u0645\u06cc \u0648 \u0642\u0627\u0628\u0644 \u0641\u0647\u0645 \u0622\u0634\u0646\u0627 \u0645\u06cc\u200c\u0634\u0648\u06cc\u062f.","breadcrumb":{"@id":"https:\/\/houshyar24.ir\/blog\/reinforcement-learning-in-llms\/#breadcrumb"},"inLanguage":"fa-IR","potentialAction":[{"@type":"ReadAction","target":["https:\/\/houshyar24.ir\/blog\/reinforcement-learning-in-llms\/"]}]},{"@type":"ImageObject","inLanguage":"fa-IR","@id":"https:\/\/houshyar24.ir\/blog\/reinforcement-learning-in-llms\/#primaryimage","url":"https:\/\/houshyar24.ir\/blog\/wp-content\/uploads\/2025\/12\/f67c7f1c-0fc5-4afd-a305-0c3086fd909c.png","contentUrl":"https:\/\/houshyar24.ir\/blog\/wp-content\/uploads\/2025\/12\/f67c7f1c-0fc5-4afd-a305-0c3086fd909c.png","width":1536,"height":1024},{"@type":"BreadcrumbList","@id":"https:\/\/houshyar24.ir\/blog\/reinforcement-learning-in-llms\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"\u062e\u0627\u0646\u0647","item":"https:\/\/houshyar24.ir\/blog\/"},{"@type":"ListItem","position":2,"name":"\u00a0\u06cc\u0627\u062f\u06af\u06cc\u0631\u06cc \u062a\u0642\u0648\u06cc\u062a\u06cc \u062f\u0631 \u0645\u062f\u0644\u200c\u0647\u0627\u06cc \u0632\u0628\u0627\u0646\u06cc: \u0686\u0631\u0627 RL \u0642\u0644\u0628 \u062a\u067e\u0646\u062f\u0647 LLM\u0647\u0627\u06cc \u0645\u062f\u0631\u0646 \u0627\u0633\u062a\u061f"}]},{"@type":"WebSite","@id":"https:\/\/houshyar24.ir\/blog\/#website","url":"https:\/\/houshyar24.ir\/blog\/","name":"\u0648\u0628\u0644\u0627\u06af \u0647\u0648\u0634\u200c\u06cc\u0627\u063124","description":"","potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/houshyar24.ir\/blog\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"fa-IR"},{"@type":"Person","@id":"https:\/\/houshyar24.ir\/blog\/#\/schema\/person\/5136eb5d0fe45d207dbfbf8f24ae5217","name":"\u0641\u0627\u0626\u0632\u0647 \u0639\u0628\u062f\u0644\u06cc \u0646\u0698\u0627\u062f","image":{"@type":"ImageObject","inLanguage":"fa-IR","@id":"https:\/\/houshyar24.ir\/blog\/#\/schema\/person\/image\/","url":"https:\/\/secure.gravatar.com\/avatar\/bbaebbb93cd807693dd89d77e905119940677df168b33f59e6b15c153249fac1?s=96&d=mm&r=g","contentUrl":"https:\/\/secure.gravatar.com\/avatar\/bbaebbb93cd807693dd89d77e905119940677df168b33f59e6b15c153249fac1?s=96&d=mm&r=g","caption":"\u0641\u0627\u0626\u0632\u0647 \u0639\u0628\u062f\u0644\u06cc \u0646\u0698\u0627\u062f"},"url":"https:\/\/houshyar24.ir\/blog\/author\/faeze\/"}]}},"_links":{"self":[{"href":"https:\/\/houshyar24.ir\/blog\/wp-json\/wp\/v2\/posts\/1236","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/houshyar24.ir\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/houshyar24.ir\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/houshyar24.ir\/blog\/wp-json\/wp\/v2\/users\/5"}],"replies":[{"embeddable":true,"href":"https:\/\/houshyar24.ir\/blog\/wp-json\/wp\/v2\/comments?post=1236"}],"version-history":[{"count":1,"href":"https:\/\/houshyar24.ir\/blog\/wp-json\/wp\/v2\/posts\/1236\/revisions"}],"predecessor-version":[{"id":1242,"href":"https:\/\/houshyar24.ir\/blog\/wp-json\/wp\/v2\/posts\/1236\/revisions\/1242"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/houshyar24.ir\/blog\/wp-json\/wp\/v2\/media\/1241"}],"wp:attachment":[{"href":"https:\/\/houshyar24.ir\/blog\/wp-json\/wp\/v2\/media?parent=1236"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/houshyar24.ir\/blog\/wp-json\/wp\/v2\/categories?post=1236"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/houshyar24.ir\/blog\/wp-json\/wp\/v2\/tags?post=1236"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}