As an example of a prompt described in primary sources, we illustrate the concept for understanding completion behavior trends when the beginning of a famous text is input.<\/p>\n\n
\n
Prompt: Four score and seven years ago our\nCompletion: fathers brought forth on this continent, a new nation, conceived in Liberty...\n<\/pre>\n<\/div>\nAdditionally, the concept is shown where the Chat Completion API sends data as an array categorized by role, as follows.<\/p>\n\n
\n
[\n {\"role\": \"system\", \"content\": \"You are a helpful assistant.\"},\n {\"role\": \"user\", \"content\": \"\u30d7\u30ed\u30f3\u30d7\u30c8\u306e\u69cb\u6210\u8981\u7d20\u306b\u3064\u3044\u3066\u8aac\u660e\u3057\u3066\u304f\u3060\u3055\u3044\u3002\"}\n]\n<\/pre>\n<\/div>\n<\/span>Results to Verify<\/span><\/h2>\n\n\nAs described in primary sources, the model outputs a response determined to have the highest probability based on training data and targets. It does not pass through a dedicated Q&A code path for questions; rather, because continuing with an answer is the most natural output, it can be confirmed that it appears to be answering as a result.<\/p>\n\n\n
<\/span>What Was Learned<\/span><\/h2>\n\n\nAll contents covered in this article are unverified on actual hardware. Within the scope confirmed from primary sources, prompts are text for communication between the user and the GPT model, and we learned that there are differences in input structure (such as arrays of dictionaries and role definitions) between the Completion API and the Chat Completion API. The primary sources also state that prompt construction has an artistic aspect and requires learning because behaviors vary across models.<\/p>\n\n\n
<\/span>Practical Notes<\/span><\/h2>\n\n\n\nThe contents of this article have not been verified for operation on actual hardware.<\/p><\/li>\n
The techniques introduced here may not be recommended for reasoning models (such as gpt-5 or the o-series).<\/p><\/li>\n
Since behaviors differ depending on the model, verification must be performed according to the actual environment and version. Please verify for yourself.<\/p><\/li>\n<\/ul>\n\n\n
<\/span>Reference Information<\/span><\/h2>\n\n\n\nsource_title: Azure OpenAI prompt engineering concepts<\/p><\/li>\n
source_url: https:\/\/learn.microsoft.com\/en-us\/azure\/foundry\/openai\/concepts\/prompt-engineering<\/p><\/li>\n<\/ul>\n","protected":false},"excerpt":{"rendered":"
This article is a technical commentary and implementation example created using AI. Although the published cod […]<\/p>\n","protected":false},"author":1,"featured_media":13321,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_jetpack_newsletter_access":"","_jetpack_dont_email_post_to_subs":false,"_jetpack_newsletter_tier_id":0,"_jetpack_memberships_contains_paywalled_content":false,"_jetpack_feature_clip_id":0,"_jetpack_memberships_contains_paid_content":false,"activitypub_content_warning":"","activitypub_content_visibility":"","activitypub_max_image_attachments":4,"activitypub_interaction_policy_quote":"anyone","activitypub_status":"federated","footnotes":"","jetpack_publicize_message":"","jetpack_publicize_feature_enabled":true,"jetpack_social_post_already_shared":true,"jetpack_social_options":{"image_generator_settings":{"template":"highway","default_image_id":0,"font":"","enabled":false},"version":2},"jetpack_post_was_ever_published":false},"categories":[970],"tags":[],"class_list":["post-13644","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-ai-machine-learning-en-post"],"aioseo_notices":[],"aioseo_head":"\n\t\t\n\t