[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-81875-en":3,"doc-seo-81875-105":31,"detail-sidebar-cat-0-en-105":93},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":21,"is_downloadable":21,"audit_status":21,"page_count":22,"language":23,"language_code":24,"site_id":25,"html_lang":24,"table_of_contents":26,"faqs":27,"seo_title":28,"seo_description":14,"update_tm":29,"read_time":30},81875,2336464648322,"Aria","https://ap-avatar.wpscdn.com/avatar/2200025388227c56fec?_k=1778556882303663488",8,"Research & Report","How to Avoid Debate: Scalable AI Safety via Doubly-Efficient Interactive Proofs","As AI capabilities expand, verifying that model outputs match human intentions becomes essential. Prior scalable oversight research has explored verification via debate, using two powerful provers to persuade a weak verifier, but this relies on unrealistic assumptions about prover parity and truthfulness. This work studies single-prover interactive proofs for AI safety, addressing limitations when computation accesses oracles. The paper develops doubly-efficient proofs and arguments for robust oracle computation and low-degree polynomial oracle access, enabling interactive verification without debate under structured or noise-tolerant oracle settings.","arXiv :2607 .0356 1v 1 [ cs .AI] 3 Jul 2026  \nHow to Avoid Debate:  \nScalable AI Safety via Doubly-Efficient Interactive Proofs  \nLiyan Chen∗ Yael Tauman Kalai† Zoe Xi‡  \nAbstract  \nAs AI models continue to develop powerful capabilities, it becomes critical that we are able to verify that their output is aligned with our intentions. A recent line of work focuses on verification via debate, a model of interactive proofs where two competing powerful provers, or AI models, debate each other to convince a weak verifier, or a human, of the correctness of their claim. However, debate assumes that the two AI models possess equal abilities and that one of them is truthful, which may not be realistic.  \nIn this work, we show how to avoid debate: we initiate the study of single-prover interactive proofs for AI safety. Prior results in single-prover interactive proofs do not immediately carry over to the AI safety setting: for example, they do not work when the computation has access to an oracle, such as to human judgment or an external database such as the web. We present doubly-efficient single-prover interactive proofs and arguments for oracle-aided computations (also known as relativizing proofs), in the settings where (1) the computation is robust, in the sense that the output does not change ifat most a small fraction of the answers to oracle queries are incorrect, or (2) the oracle is a low-degree polynomial. These results suggest that interactive verification is possible even without debate, under structured or noise-tolerant oracle access.  \n∗[MIT.](MIT. cliyan@mit.edu)[ cliyan@mit.edu](MIT. cliyan@mit.edu).†[MIT.](MIT. tauman@mit.edu)[ tauman@mit.edu](MIT. tauman@mit.edu).‡[MIT.](MIT. zoexi@mit.edu)[ zoexi@mit.edu](MIT. zoexi@mit.edu).  \nContents  \n1 Introduction 3  \n1.1 Our Results .......................................... 4  \n1.2 Related Work ......................................... 5  \n2 Technical Overview 6  \n2.1 Proof system for robust oracle computation ........................ 6  \n2.2 Argument system for robust oracle computation ...................... 8  \n2.3 Argument system for computation with access to a low-degree oracle .......... 8  \n3 Preliminaries 9  \n3.1 Oracle-Aided Computation and Robustness ........................ 9  \n3.2 Succinct Descriptions of Sets and Functions ........................ 10  \n3.3 Interactive Proofs ....................................... 11  \n3.4 Interactive Proofs of Proximity ............................... 12  \n3.5 Unique-Decoding Checksums ................................ 13  \n3.6 Interactive Arguments and Arguments of Knowledge ................... 14  \n3.7 Hash Trees .......................................... 15  \n3.8 Polynomial Commitments .................................. 16  \n4 Doubly-Efficient Interactive Proofs for Robust Oracle Circuits 17  \n4.1 A Proof System for Circuits Making Nonadaptive Oracle Queries ............ 17  \n4.2 A Proof System for Circuits Making Adaptive Oracle Queries .............. 21  \n4.3 An Argument System for Circuits Making Adaptive Oracle Queries ........... 27  \n5 A Doubly-Efficient Interactive Argument for Circuits with a Low-Degree Oracle 29  \n6 Conclusion 32  \n7 Acknowledgments 33  \n1 Introduction  \nAs machine learning models become increasingly capable, they are being trained to perform complex tasks that would be prohibitively expensive for a human to verify, leading to critical safety concerns. How can we efficiently supervise the training of an AI model to perform some complex task in a way that aligns with our intentions, when we may only have an obscure understanding of the task ourselves?  \nAs a concrete example, consider the scenario of training a large language model (LLM) to write along high-stakes document, such as a legal contract.1 To generate a training label on a contract produced by the LLM, it is necessary to verify that every passage of the contract is correct, where correctness here is dictated by human judgment. However, it w","cbCaiiEAYgEROa5N","https://ap.wps.com/l/cbCaiiEAYgEROa5N","pdf",674093,6,1,35,"English","en",105,"# Introduction\n## Our Results\n## Related Work\n# Technical Overview\n## Proof System for Robust Oracle Computation\n## Argument System for Robust Oracle Computation\n## Argument System for Low-Degree Oracle Access\n# Doubly-Efficient Interactive Proofs and Arguments\n## Robust Oracle Circuits (Nonadaptive and Adaptive Queries)\n## Low-Degree Oracle Argument\n# Conclusion\n# Acknowledgments","[{\"question\":\"Why does debate-based verification become problematic for AI safety?\",\"answer\":\"Debate-based interactive proofs assume two provers have equal abilities and that at least one is truthful, which may not reflect realistic AI settings.\"},{\"question\":\"What is the core contribution of this work?\",\"answer\":\"The work initiates single-prover interactive proofs for AI safety and shows how verification can be performed without requiring debate.\"},{\"question\":\"Under what oracle models do the proposed doubly-efficient results work?\",\"answer\":\"The results apply to oracle-aided computations with either robust computation (small fractions of incorrect oracle answers) or when the oracle is a low-degree polynomial.\"}]","How to Avoid Debate: Scalable AI Safety via Doubly-Efficient Interactive Proofs | PDF",1784176800,88,{"code":4,"msg":32,"data":33},"ok",{"site_id":25,"language":24,"slug":34,"title":13,"keywords":35,"description":14,"schema_data":36,"social_meta":88,"head_meta":90,"extra_data":92,"updated_unix":29},"how-to-avoid-debate-scalable-ai-safety-via-doubly-efficient-interactive-proofs","",{"@graph":37,"@context":87},[38,55,70],{"@type":39,"itemListElement":40},"BreadcrumbList",[41,45,49,52],{"item":42,"name":43,"@type":44,"position":21},"https://docshare.wps.com","Home","ListItem",{"item":46,"name":47,"@type":44,"position":48},"https://docshare.wps.com/document/","Document",2,{"item":50,"name":12,"@type":44,"position":51},"https://docshare.wps.com/document/research-report/",3,{"item":53,"name":13,"@type":44,"position":54},"https://docshare.wps.com/document/how-to-avoid-debate-scalable-ai-safety-via-doubly-efficient-interactive-proofs/81875/",4,{"url":53,"name":13,"@type":56,"author":57,"headline":13,"publisher":59,"fileFormat":62,"inLanguage":24,"description":14,"dateModified":63,"datePublished":64,"encodingFormat":62,"isAccessibleForFree":65,"interactionStatistic":66},"DigitalDocument",{"name":9,"@type":58},"Person",{"url":42,"name":60,"@type":61},"DocShare","Organization","application/pdf","2026-08-03","2026-07-16",true,{"@type":67,"interactionType":68,"userInteractionCount":20},"InteractionCounter",{"@type":69},"ViewAction",{"@type":71,"mainEntity":72},"FAQPage",[73,79,83],{"name":74,"@type":75,"acceptedAnswer":76},"Why does debate-based verification become problematic for AI safety?","Question",{"text":77,"@type":78},"Debate-based interactive proofs assume two provers have equal abilities and that at least one is truthful, which may not reflect realistic AI settings.","Answer",{"name":80,"@type":75,"acceptedAnswer":81},"What is the core contribution of this work?",{"text":82,"@type":78},"The work initiates single-prover interactive proofs for AI safety and shows how verification can be performed without requiring debate.",{"name":84,"@type":75,"acceptedAnswer":85},"Under what oracle models do the proposed doubly-efficient results work?",{"text":86,"@type":78},"The results apply to oracle-aided computations with either robust computation (small fractions of incorrect oracle answers) or when the oracle is a low-degree polynomial.","https://schema.org",{"og:url":53,"og:type":89,"og:title":13,"og:site_name":60,"og:description":14},"article",{"robots":91,"canonical":53},"index,follow",{"doc_id":7,"site_id":25},{"code":4,"msg":5,"data":94},[95,99,103,107,112,116,121,124,129,132,136],{"id":21,"doc_module":4,"doc_module_name":47,"category_name":96,"show_sort_weight":97,"slug":98},"Story & Novel",90,"story-novel",{"id":48,"doc_module":4,"doc_module_name":47,"category_name":100,"show_sort_weight":101,"slug":102},"Literature",80,"literature",{"id":54,"doc_module":4,"doc_module_name":47,"category_name":104,"show_sort_weight":105,"slug":106},"Exam",70,"exam",{"id":108,"doc_module":4,"doc_module_name":47,"category_name":109,"show_sort_weight":110,"slug":111},5,"Comic",60,"comic",{"id":20,"doc_module":4,"doc_module_name":47,"category_name":113,"show_sort_weight":114,"slug":115},"Technology",50,"technology",{"id":117,"doc_module":4,"doc_module_name":47,"category_name":118,"show_sort_weight":119,"slug":120},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":47,"category_name":12,"show_sort_weight":122,"slug":123},30,"research-report",{"id":125,"doc_module":4,"doc_module_name":47,"category_name":126,"show_sort_weight":127,"slug":128},9,"Religion & Spirituality",20,"religion-spirituality",{"id":127,"doc_module":4,"doc_module_name":47,"category_name":130,"show_sort_weight":127,"slug":131},"World Cup","world-cup",{"id":133,"doc_module":4,"doc_module_name":47,"category_name":134,"show_sort_weight":133,"slug":135},10,"Lifestyle","lifestyle",{"id":137,"doc_module":4,"doc_module_name":47,"category_name":138,"show_sort_weight":108,"slug":139},19,"General","general"]