{"id":1083801,"date":"2024-09-10T08:56:57","date_gmt":"2024-09-10T15:56:57","guid":{"rendered":"https:\/\/www.microsoft.com\/en-us\/research\/?post_type=msr-project&p=1083801"},"modified":"2024-09-10T08:57:01","modified_gmt":"2024-09-10T15:57:01","slug":"auto-tuning-of-copilot-prompts","status":"publish","type":"msr-project","link":"https:\/\/www.microsoft.com\/en-us\/research\/project\/auto-tuning-of-copilot-prompts\/","title":{"rendered":"Auto-Tuning of Copilot Prompts"},"content":{"rendered":"
\n\t
\n\t\t
\n\t\t\t\"Microsoft\t\t<\/div>\n\t\t\n\t\t
\n\t\t\t\n\t\t\t
\n\t\t\t\t\n\t\t\t\t
\n\t\t\t\t\t\n\t\t\t\t\t
\n\t\t\t\t\t\t
\n\t\t\t\t\t\t\t\n\t\t\t\t\t\t\t\n\n

Auto-Tuning of Copilot Prompts<\/h1>\n\n\n\n

AI-agent-based tool for tuning and optimizing system prompts, and particularly those of Microsoft Copilot offerings. It is used to replace manual adaptation of prompts — requiring extensive human labor — in scenarios such as when the underlying language models are updated.<\/p>\n\n\t\t\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t<\/div>\n\t\t<\/div>\n\t<\/div>\n<\/section>\n\n\n\n\n\n

Microsoft Copilot offerings use state-of-the-art generative AI technology to empower people and help them fulfill their tasks. This technology includes prompt-based methods utilizing some of the largest language models to date. Crafting system prompts is difficult and time-consuming and requires considerable human labor. It is needed not only when updating the experience or when introducing new capabilities, but also with each incremental update of the underlying language models. In this project, we build an effective pipeline of AI agents for the automated tuning and optimization of the Copilot system prompts. AI is used to evaluate prompts and prompt-generated content, reason, and generate new and improved prompts.<\/p>\n\n\n","protected":false},"excerpt":{"rendered":"

AI-agent-based tool for tuning and optimizing system prompts, and particularly those of Microsoft Copilot offerings. It is used to replace manual adaptation of prompts — requiring extensive human labor — in scenarios such as when the underlying language models are updated. Microsoft Copilot offerings use state-of-the-art generative AI technology to empower people and help them […]<\/p>\n","protected":false},"featured_media":1083837,"template":"","meta":{"msr-url-field":"","msr-podcast-episode":"","msrModifiedDate":"","msrModifiedDateEnabled":false,"ep_exclude_from_search":false,"_classifai_error":"","footnotes":""},"research-area":[13556,13545],"msr-locale":[268875],"msr-impact-theme":[],"msr-pillar":[],"class_list":["post-1083801","msr-project","type-msr-project","status-publish","has-post-thumbnail","hentry","msr-research-area-artificial-intelligence","msr-research-area-human-language-technologies","msr-locale-en_us","msr-archive-status-active"],"msr_project_start":"","related-publications":[],"related-downloads":[],"related-videos":[],"related-groups":[],"related-events":[],"related-opportunities":[],"related-posts":[],"related-articles":[],"tab-content":[],"slides":[],"related-researchers":[{"type":"user_nicename","display_name":"Amir Kantor","user_id":40153,"people_section":"Related people","alias":"amkantor"},{"type":"user_nicename","display_name":"Eyal Kolman","user_id":41006,"people_section":"Related people","alias":"eyalko"},{"type":"guest","display_name":"Coral Malachi","user_id":1083819,"people_section":"Related people","alias":""},{"type":"guest","display_name":"Shlomi Maliah","user_id":1083816,"people_section":"Related people","alias":""},{"type":"guest","display_name":"Ido Priness","user_id":1083813,"people_section":"Related people","alias":""}],"msr_research_lab":[],"msr_impact_theme":[],"_links":{"self":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-project\/1083801"}],"collection":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-project"}],"about":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/types\/msr-project"}],"version-history":[{"count":5,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-project\/1083801\/revisions"}],"predecessor-version":[{"id":1083924,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-project\/1083801\/revisions\/1083924"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media\/1083837"}],"wp:attachment":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media?parent=1083801"}],"wp:term":[{"taxonomy":"msr-research-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/research-area?post=1083801"},{"taxonomy":"msr-locale","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-locale?post=1083801"},{"taxonomy":"msr-impact-theme","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-impact-theme?post=1083801"},{"taxonomy":"msr-pillar","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-pillar?post=1083801"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}