{"id":1155409,"date":"2025-12-04T04:12:24","date_gmt":"2025-12-04T12:12:24","guid":{"rendered":"https:\/\/www.microsoft.com\/en-us\/research\/?post_type=msr-project&p=1155409"},"modified":"2025-12-17T01:13:44","modified_gmt":"2025-12-17T09:13:44","slug":"agentic-verifiers-provably-safe-test-time-scaling-for-reasoning-models","status":"publish","type":"msr-project","link":"https:\/\/www.microsoft.com\/en-us\/research\/project\/agentic-verifiers-provably-safe-test-time-scaling-for-reasoning-models\/","title":{"rendered":"Agentic Verifiers: Provably Safe Test-time scaling for Reasoning Models"},"content":{"rendered":"
\n\t
\n\t\t
\n\t\t\t\"background\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

Agentic Verifiers: Provably Safe Test-time scaling for Reasoning Models<\/h1>\n\n\n\n

<\/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

This project introduces a novel architecture for agentic AI systems that ensures accuracy, efficiency, and safety during reasoning. It addresses two key challenges\u2014lack of steerability and absence of verifiable guarantees\u2014by developing verifiers that can interject at any point in a model\u2019s generation process. An auxiliary monitor model evaluates each reasoning step against predefined properties, rolling back and correcting errors in real time. The research spans commonsense, medical, and legal reasoning, aiming to deliver publicly available verifier models and an open-source platform for integrating verification into agentic systems, paving the way for trustworthy AI in high-stakes domains.

This research is conducted via The Agentic AI Research and Innovation <\/a>(AARI) Initiative which focuses on the next frontier of agentic systems through Grand Challenges<\/em> with the academic community and Microsoft Research.<\/p>\n\n\n","protected":false},"excerpt":{"rendered":"

This project introduces a novel architecture for agentic AI systems that ensures accuracy, efficiency, and safety during reasoning. It addresses two key challenges\u2014lack of steerability and absence of verifiable guarantees\u2014by developing verifiers that can interject at any point in a model\u2019s generation process. An auxiliary monitor model evaluates each reasoning step against predefined properties, rolling […]<\/p>\n","protected":false},"featured_media":1155701,"template":"","meta":{"msr-url-field":"","msr-podcast-episode":"","msrModifiedDate":"","msrModifiedDateEnabled":false,"ep_exclude_from_search":false,"_classifai_error":"","footnotes":""},"research-area":[13556],"msr-locale":[268875],"msr-impact-theme":[],"msr-pillar":[],"class_list":["post-1155409","msr-project","type-msr-project","status-publish","has-post-thumbnail","hentry","msr-research-area-artificial-intelligence","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":"guest","display_name":"Somak Aditya","user_id":1157209,"people_section":"Section name 0","alias":""},{"type":"guest","display_name":"Sourangshu Bhattacharya","user_id":1157211,"people_section":"Section name 0","alias":""},{"type":"user_nicename","display_name":"Vineeth N Balasubramanian","user_id":44019,"people_section":"Section name 0","alias":"vineethn"},{"type":"user_nicename","display_name":"Nagarajan Natarajan","user_id":37311,"people_section":"Section name 0","alias":"nagarajn"},{"type":"guest","display_name":"Uma Satya Ranjan","user_id":1157216,"people_section":"Section name 0","alias":""},{"type":"user_nicename","display_name":"Amit Sharma","user_id":30997,"people_section":"Section name 0","alias":"amshar"}],"msr_research_lab":[],"msr_impact_theme":[],"_links":{"self":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-project\/1155409","targetHints":{"allow":["GET"]}}],"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":11,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-project\/1155409\/revisions"}],"predecessor-version":[{"id":1158811,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-project\/1155409\/revisions\/1158811"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media\/1155701"}],"wp:attachment":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media?parent=1155409"}],"wp:term":[{"taxonomy":"msr-research-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/research-area?post=1155409"},{"taxonomy":"msr-locale","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-locale?post=1155409"},{"taxonomy":"msr-impact-theme","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-impact-theme?post=1155409"},{"taxonomy":"msr-pillar","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-pillar?post=1155409"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}