{"id":1156922,"date":"2025-12-01T19:43:47","date_gmt":"2025-12-02T03:43:47","guid":{"rendered":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/?post_type=msr-project&#038;p=1156922"},"modified":"2025-12-01T19:43:50","modified_gmt":"2025-12-02T03:43:50","slug":"companionx","status":"publish","type":"msr-project","link":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/project\/companionx\/","title":{"rendered":"CompanionX"},"content":{"rendered":"<section class=\"mb-3 moray-highlight\">\n\t<div class=\"card-img-overlay mx-lg-0\">\n\t\t<div class=\"card-background  has-background-gable-green card-background--full-bleed\">\n\t\t\t\t\t<\/div>\n\t\t<!-- Foreground -->\n\t\t<div class=\"card-foreground d-flex mt-md-n5 my-lg-5 px-g px-lg-0\">\n\t\t\t<!-- Container -->\n\t\t\t<div class=\"container d-flex mt-md-n5 my-lg-5 \">\n\t\t\t\t<!-- Card wrapper -->\n\t\t\t\t<div class=\"w-100 w-lg-col-5\">\n\t\t\t\t\t<!-- Card -->\n\t\t\t\t\t<div class=\"card material-md-card py-5 px-md-5\">\n\t\t\t\t\t\t<div class=\"card-body \">\n\t\t\t\t\t\t\t\n\t\t\t\t\t\t\t\n\n<h1 class=\"wp-block-heading\" id=\"companionx\">CompanionX<\/h1>\n\n\n\n<p><\/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<p>CompanionX&nbsp;moves AI beyond narrow task solving toward human-like collaboration. While modern models excel at mathematics and coding, qualities such as social intelligence, empathy, and benevolence remain underexplored.&nbsp;CompanionX&nbsp;develops agents that understand, respond to, and cooperate with people authentically and proactively, forming personalized minds and communicating with emotional competence. These digital companions will support social interaction, customer service, digital coworkers, personal coaching,&nbsp;AI NPCs,&nbsp;and large-scale societal simulations\u2014deepening trust, engagement, and real-world value.<\/p>\n\n\n\n<p>To realize this vision,&nbsp;CompanionX&nbsp;introduces a systematic tech stack for evaluating and enhancing human-like (anthropomorphic) intelligence: foundational social-cognitive capabilities and rigorous evaluations; an agentic framework for human-like task execution and collaboration; social simulation for self-play, safe experimentation, and continual improvement; and Companion Guard, a safety and ethics layer ensuring prosocial behavior, transparency, privacy, and regulatory readiness. Through interdisciplinary research and comprehensive evaluation,&nbsp;CompanionX&nbsp;advances trustworthy, collaborative, and socially adept AI that bridges technology and genuine human experience.<\/p>\n\n\n\n<p><strong>Representative Publications:<\/strong>&nbsp;<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" rel=\"noopener noreferrer\" target=\"_blank\" href=\"https:\/\/aclanthology.org\/2025.acl-long.1394.pdf\">Pretraining context compressor for large language models with embedding-based memory. ACL 2025.<span class=\"sr-only\"> (opens in new tab)<\/span><\/a><\/li>\n\n\n\n<li><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" rel=\"noopener noreferrer\" target=\"_blank\" href=\"https:\/\/aclanthology.org\/2025.acl-long.1594.pdf\">SocialCC: Interactive Evaluation for Cultural Competence in Language Agents. ACL 2025.&nbsp;<span class=\"sr-only\"> (opens in new tab)<\/span><\/a><\/li>\n\n\n\n<li><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" rel=\"noopener noreferrer\" target=\"_blank\" href=\"https:\/\/arxiv.org\/pdf\/2506.13065\">MotiveBench: How Far Are We&nbsp;From&nbsp;Human-Like Motivational Reasoning in Large Language Models? ACL 2025 Findings.&nbsp;&nbsp;<span class=\"sr-only\"> (opens in new tab)<\/span><\/a><\/li>\n\n\n\n<li><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" rel=\"noopener noreferrer\" target=\"_blank\" href=\"https:\/\/arxiv.org\/pdf\/2506.13464\">Unveiling&nbsp;the Learning Mind of Language Models: A Cognitive Framework and Empirical Study.&nbsp;NeurIPS&nbsp;2025.&nbsp;<span class=\"sr-only\"> (opens in new tab)<\/span><\/a><\/li>\n\n\n\n<li><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" rel=\"noopener noreferrer\" target=\"_blank\" href=\"https:\/\/aclanthology.org\/2025.naacl-long.323.pdf\">CharacterBox: Evaluating the Role-Playing Capabilities of LLMs in Text-based Virtual Worlds. NAACL 2025.<span class=\"sr-only\"> (opens in new tab)<\/span><\/a><\/li>\n\n\n\n<li>HumanLLM: Towards Personalized Understanding and Simulation of Human Nature. KDD 2026. To appear.&nbsp;<\/li>\n\n\n\n<li><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" rel=\"noopener noreferrer\" target=\"_blank\" href=\"https:\/\/arxiv.org\/pdf\/2509.10127\">Population-Aligned Persona Generation for LLM-based Social Simulation. Preprint.&nbsp;<span class=\"sr-only\"> (opens in new tab)<\/span><\/a><\/li>\n<\/ul>\n\n\n\n<p><strong>Open-source contributions:<\/strong>&nbsp;<\/p>\n\n\n\n<p><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" rel=\"noopener noreferrer\" target=\"_blank\" href=\"https:\/\/github.com\/microsoft\/AnthropomorphicIntelligence\">https:\/\/github.com\/microsoft\/AnthropomorphicIntelligence<span class=\"sr-only\"> (opens in new tab)<\/span><\/a><\/p>\n\n\n\n<p><strong>Other achievements:<\/strong>&nbsp;<\/p>\n\n\n\n<p>Our&nbsp;v-team (Yitian Huang, Yuxuan Lei, Jianxun Lian) won one first place, one second place, and two third place&nbsp;awards&nbsp;in&nbsp;<a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/www.aicrowd.com\/challenges\/commonsense-persona-grounded-dialogue-challenge-2025\" target=\"_blank\" rel=\"noopener noreferrer\">CPDC Challenge 2025<span class=\"sr-only\"> (opens in new tab)<\/span><\/a>, a competition&nbsp;focused on&nbsp;AI NPC.&nbsp;<a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/arxiv.org\/pdf\/2511.20200\" target=\"_blank\" rel=\"noopener noreferrer\">Interactive AI NPCs Powered by LLMs: Technical&nbsp;Report for the CPDC Challenge 2025<span class=\"sr-only\"> (opens in new tab)<\/span><\/a>.&nbsp;&nbsp;<\/p>\n\n\n\n\n\n<p><\/p>\n\n\n","protected":false},"excerpt":{"rendered":"<p>CompanionX&nbsp;moves AI beyond narrow task solving toward human-like collaboration. While modern models excel at mathematics and coding, qualities such as social intelligence, empathy, and benevolence remain underexplored.&nbsp;CompanionX&nbsp;develops agents that understand, respond to, and cooperate with people authentically and proactively, forming personalized minds and communicating with emotional competence. These digital companions will support social interaction, customer [&hellip;]<\/p>\n","protected":false},"featured_media":0,"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-1156922","msr-project","type-msr-project","status-publish","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":"user_nicename","display_name":"Jianxun Lian","user_id":38470,"people_section":"Section name 0","alias":"jialia"},{"type":"user_nicename","display_name":"Haotian Li","user_id":43593,"people_section":"Section name 0","alias":"haotianli"},{"type":"user_nicename","display_name":"Liming Jiang","user_id":43996,"people_section":"Section name 0","alias":"jiangliming"},{"type":"user_nicename","display_name":"Xing Xie","user_id":34906,"people_section":"Section name 0","alias":"xingx"},{"type":"user_nicename","display_name":"Beibei Shi","user_id":42162,"people_section":"Section name 0","alias":"besh"}],"msr_research_lab":[199560,1012650],"msr_impact_theme":[],"_links":{"self":[{"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/msr-project\/1156922","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/msr-project"}],"about":[{"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/types\/msr-project"}],"version-history":[{"count":6,"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/msr-project\/1156922\/revisions"}],"predecessor-version":[{"id":1157182,"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/msr-project\/1156922\/revisions\/1157182"}],"wp:attachment":[{"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/media?parent=1156922"}],"wp:term":[{"taxonomy":"msr-research-area","embeddable":true,"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/research-area?post=1156922"},{"taxonomy":"msr-locale","embeddable":true,"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/msr-locale?post=1156922"},{"taxonomy":"msr-impact-theme","embeddable":true,"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/msr-impact-theme?post=1156922"},{"taxonomy":"msr-pillar","embeddable":true,"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/msr-pillar?post=1156922"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}