[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"detail-sidebar-cat-0-en-105":3,"doc-seo-290819-105":59,"doc-detail-290819-en":130},{"code":4,"msg":5,"data":6},0,"success",[7,13,18,23,28,33,38,43,48,51,55],{"id":8,"doc_module":4,"doc_module_name":9,"category_name":10,"show_sort_weight":11,"slug":12},1,"Document","Story & Novel",90,"story-novel",{"id":14,"doc_module":4,"doc_module_name":9,"category_name":15,"show_sort_weight":16,"slug":17},2,"Literature",80,"literature",{"id":19,"doc_module":4,"doc_module_name":9,"category_name":20,"show_sort_weight":21,"slug":22},4,"Exam",70,"exam",{"id":24,"doc_module":4,"doc_module_name":9,"category_name":25,"show_sort_weight":26,"slug":27},5,"Comic",60,"comic",{"id":29,"doc_module":4,"doc_module_name":9,"category_name":30,"show_sort_weight":31,"slug":32},6,"Technology",50,"technology",{"id":34,"doc_module":4,"doc_module_name":9,"category_name":35,"show_sort_weight":36,"slug":37},7,"Healthcare",40,"healthcare",{"id":39,"doc_module":4,"doc_module_name":9,"category_name":40,"show_sort_weight":41,"slug":42},8,"Research & Report",30,"research-report",{"id":44,"doc_module":4,"doc_module_name":9,"category_name":45,"show_sort_weight":46,"slug":47},9,"Religion & Spirituality",20,"religion-spirituality",{"id":46,"doc_module":4,"doc_module_name":9,"category_name":49,"show_sort_weight":46,"slug":50},"World Cup","world-cup",{"id":52,"doc_module":4,"doc_module_name":9,"category_name":53,"show_sort_weight":52,"slug":54},10,"Lifestyle","lifestyle",{"id":56,"doc_module":4,"doc_module_name":9,"category_name":57,"show_sort_weight":24,"slug":58},19,"General","general",{"code":4,"msg":60,"data":61},"ok",{"site_id":62,"language":63,"slug":64,"title":65,"keywords":66,"description":67,"schema_data":68,"social_meta":123,"head_meta":125,"extra_data":127,"updated_unix":129},105,"en","responsiveness-practical-safety-in-machine-learning-under-interaction","Responsiveness: Practical Safety in Machine Learning Under Interaction","","Responsiveness: Practical Safety in Machine Learning Under Interaction develops a framework for analyzing and validating model safety in interactive settings, where system outputs influence subsequent user behavior and decision dynamics. It introduces feature responsiveness scores to produce safe explanations, and it proposes responsiveness verification methods to validate safety using statistical certification and stress testing. The thesis includes algorithms and experimental and demonstration studies, including case analyses for recidivism prediction preclusion, content moderation gaming, and large language model safety assessment.",{"@graph":69,"@context":122},[70,84,105],{"@type":71,"itemListElement":72},"BreadcrumbList",[73,77,79,82],{"item":74,"name":75,"@type":76,"position":8},"https://docshare.wps.com","Home","ListItem",{"item":78,"name":9,"@type":76,"position":14},"https://docshare.wps.com/document/",{"item":80,"name":40,"@type":76,"position":81},"https://docshare.wps.com/document/research-report/",3,{"item":83,"name":65,"@type":76,"position":19},"https://docshare.wps.com/document/responsiveness-practical-safety-in-machine-learning-under-interaction/290819/",{"url":83,"name":65,"@type":85,"image":86,"author":91,"headline":65,"publisher":94,"fileFormat":97,"inLanguage":63,"description":67,"dateModified":98,"datePublished":99,"encodingFormat":97,"isAccessibleForFree":100,"interactionStatistic":101},"DigitalDocument",{"url":87,"@type":88,"width":89,"height":90},"https://docshare.wps.com/thumbnails/responsiveness-practical-safety-in-machine-learning-under-interaction/290819.png","ImageObject",300,407,{"name":92,"@type":93},"Skyler","Person",{"url":74,"name":95,"@type":96},"DocShare","Organization","application/pdf","2026-10-17","2026-09-17",true,{"@type":102,"interactionType":103,"userInteractionCount":52},"InteractionCounter",{"@type":104},"ViewAction",{"@type":106,"mainEntity":107},"FAQPage",[108,114,118],{"name":109,"@type":110,"acceptedAnswer":111},"What problem does the thesis address?","Question",{"text":112,"@type":113},"It addresses practical safety in machine learning when models are deployed under interaction, where outputs can change downstream behavior and decisions.","Answer",{"name":115,"@type":110,"acceptedAnswer":116},"How does the thesis generate safe explanations?",{"text":117,"@type":113},"It introduces feature responsiveness scores to quantify how model features respond under interactive settings, supporting safe explanation mechanisms.",{"name":119,"@type":110,"acceptedAnswer":120},"How is model safety validated in the interactive setting?",{"text":121,"@type":113},"It proposes responsiveness verification, combining statistical certification with stress testing of interaction model assumptions.","https://schema.org",{"og:url":83,"og:type":124,"og:title":65,"og:site_name":95,"og:description":67},"article",{"robots":126,"canonical":83},"index,follow",{"doc_id":128,"site_id":62},290819,1790027381,{"code":4,"msg":5,"data":131},{"doc_id":128,"user_id":132,"nickname":92,"user_avatar":133,"doc_module":4,"category_id":39,"category_name":40,"doc_title":65,"doc_description":67,"doc_content":134,"file_id":135,"file_url":136,"file_type":137,"file_size":138,"view_count":52,"is_deleted":4,"is_public":8,"is_downloadable":8,"audit_status":8,"page_count":139,"language":140,"language_code":63,"site_id":62,"html_lang":63,"table_of_contents":141,"faqs":142,"seo_title":143,"seo_description":67,"update_tm":144,"read_time":145,"preview_page_count":4},2336464648746,"https://ap-avatar.wpscdn.com/davatar_276721f389ce27ea32af1340a28f341c","UC San Diego  \nUC San Diego Electronic Theses and Dissertations  \nTitle  \nResponsiveness: Practical Safety in Machine Learning Under Interaction  \nPermalink  \n[https://escholarship.org/uc/item/24n4p20m](https://escholarship.org/uc/item/24n4p20m)  \nISBN  \n9798197806697  \nAuthor  \nCheon, Harry  \nPublication Date  \n2026-06-26  \nPeer reviewed|Thesis/dissertation  \n[eScholarship.org](eScholarship.org) Powered by the California Digital Library  \nUniversity of California  \nUNIVERSITY OF CALIFORNIA SAN DIEGO  \nResponsiveness: Practical Safety in Machine Learning Under Interaction  \nA thesis submitted in partial satisfaction of the  \nrequirements for the degree Master of Science  \nin  \nComputer Science  \nby  \nHarry Cheon  \nCommittee in charge:  \nProfessor Berk Ustun, Chair  \nProfessor Julian McAuley, Co-Chair  \nProfessor Loris D’Antoni  \nCopyright Harry Cheon, 2026 All rights reserved.  \nThe thesis of Harry Cheon is approved, and it is acceptable in quality and form for publication on microfilm and electronically.  \nUniversity of California San Diego  \n2026  \nTo God, my family, and friends.  \niv  \nTABLE OF CONTENTS  \nThesis Approval Page ......................................................... iii  \nDedication .................................................................. iv  \nTable of Contents ............................................................ v  \nList of Figures ............................................................... viii  \nList of Tables ................................................................ x  \nAcknowledgements ........................................................... xii  \nAbstract of the Thesis ......................................................... xiii  \nChapter 1 Introduction ..................................................... 1  \n1.1 Contributions ........................................................ 1  \n1.2 Thesis Organization .................................................. 2  \nChapter 2 Overarching Framework: Interaction Model .......................... 4  \nChapter 3 Safe Explanations Under Interactions: Feature Responsiveness Scores .... 6  \n3.1 Introduction ......................................................... 6  \n3.1.1 Related Work ................................................. 8  \n3.2 Problem Statement ................................................... 9  \n3.3 Measuring Feature Responsiveness ...................................... 12  \n3.4 Computing Scores with Reachable Sets .................................. 14  \n3.5 Experiments ......................................................... 17  \n3.6 On the Limits of Feature Highlighting for Recourse ........................ 21  \n3.7 Concluding Remarks .................................................. 25  \nChapter 4 Validating Model Safety Under Interactions: Responsiveness Verification . 27  \n4.1 Introduction ......................................................... 27  \n4.1.1 Related Work ................................................. 30  \n4.2 Validating Model Safety with Responsiveness ............................. 30  \n4.2.1 Statistical Certification ......................................... 31  \n4.2.2 Stress Testing Interaction Model Assumptions ...................... 33  \n4.3 Algorithms .......................................................... 34  \n4.4 Demonstrations ...................................................... 38  \n4.4.1 Preclusion in Recidivism Prediction .............................. 38  \n4.4.2 Gaming in Content Moderation .................................. 40  \n4.4.3 Evaluating Large Language Model Safety under Responsiveness ...... 41  \n4.5 Concluding Remarks .................................................. 43  \nAppendix A Supplementary Material for Chapter 3 ............................... 44  \nA.1 Supplementary Material for Section 3.4 .................................. 44  \nA.1.1 Implementation Details for Reachable Set Enumeration .............. 44  \nA.1.2 Implementation Details for Reacha","cbCaiaZqb2XS6anK","https://ap.wps.com/l/cbCaiaZqb2XS6anK","pdf",3002392,133,"English","# Thesis Approval Page\n# Dedication\n# List of Figures\n# List of Tables\n# Acknowledgements\n# Abstract of the Thesis\n# Chapter 1 Introduction\n## Contributions\n## Thesis Organization\n# Chapter 2 Overarching Framework: Interaction Model\n# Chapter 3 Safe Explanations Under Interactions: Feature Responsiveness Scores\n## Introduction\n## Related Work\n## Problem Statement\n## Measuring Feature Responsiveness\n## Computing Scores with Reachable Sets\n## Experiments\n## On the Limits of Feature Highlighting for Recourse\n## Concluding Remarks\n# Chapter 4 Validating Model Safety Under Interactions: Responsiveness Verification\n## Introduction\n## Related Work\n## Validating Model Safety with Responsiveness\n## Statistical Certification\n## Stress Testing Interaction Model Assumptions\n## Algorithms\n## Demonstrations\n## Preclusion in Recidivism Prediction\n## Gaming in Content Moderation\n## Evaluating Large Language Model Safety under Responsiveness\n## Concluding Remarks\n# Appendix A Supplementary Material for Chapter 3\n## Supplementary Material for Section 3.4\n## Supplementary Experiment Details\n## Supplementary Experiment Results\n## Supplementary Case Study Details\n# Appendix B Supplementary Material for Chapter 4\n## Formal Procedure for Responsiveness Validation Tasks\n## Description of the Mixed Integer Program\n## Supplementary Material for Section 4.3","[{\"question\":\"What problem does the thesis address?\",\"answer\":\"It addresses practical safety in machine learning when models are deployed under interaction, where outputs can change downstream behavior and decisions.\"},{\"question\":\"How does the thesis generate safe explanations?\",\"answer\":\"It introduces feature responsiveness scores to quantify how model features respond under interactive settings, supporting safe explanation mechanisms.\"},{\"question\":\"How is model safety validated in the interactive setting?\",\"answer\":\"It proposes responsiveness verification, combining statistical certification with stress testing of interaction model assumptions.\"}]","Responsiveness: Practical Safety in Machine Learning Under Interaction | PDF",1789642647,335]