[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"detail-sidebar-cat-0-en-105":3,"doc-seo-152660-105":59,"doc-detail-152660-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","psycho-personality-characterization-of-artificial-intelligence","PSYCHO - PerSonalitY CHaracterizatiOn of artificial intelligence","","Recent advances in AI driven by deep learning have transformed everyday life, while also raising fears about unintended consequences. This work proposes a framework to examine the “brains” of modern robot models by adapting human psychology research. The method applies the Rorschach inkblot paradigm to deep networks, using nearest-neighbor matching for candidate responses and reporting quantitative and qualitative personality-like behaviors.",{"@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/psycho-personality-characterization-of-artificial-intelligence/152660/",{"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/psycho-personality-characterization-of-artificial-intelligence/152660.png","ImageObject",300,407,{"name":92,"@type":93},"Sophia Brooks","Person",{"url":74,"name":95,"@type":96},"DocShare","Organization","application/pdf","2026-11-05","2026-08-27",true,{"@type":102,"interactionType":103,"userInteractionCount":39},"InteractionCounter",{"@type":104},"ViewAction",{"@type":106,"mainEntity":107},"FAQPage",[108,114,118],{"name":109,"@type":110,"acceptedAnswer":111},"What problem does PSYCHO aim to address regarding AI advancements?","Question",{"text":112,"@type":113},"It addresses concerns that powerful AI systems could behave in harmful ways, arguing that the key need is mental/behavioral interpretability rather than simply adding more safeguards of the same kind.","Answer",{"name":115,"@type":110,"acceptedAnswer":116},"How does the method use the Rorschach inkblot test to analyze AI models?",{"text":117,"@type":113},"It treats each inkblot as an input and selects the best matching response option by finding nearest neighbors of the model’s representations corresponding to each candidate response.",{"name":119,"@type":110,"acceptedAnswer":120},"Why does the approach rely on image search instead of manually collecting visuals?",{"text":121,"@type":113},"It aims to reduce bias that can arise from biased or non-representative image collection by directly querying an image search engine for images associated with each potential response.","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},152660,1787865479,{"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":39,"is_deleted":4,"is_public":8,"is_downloadable":8,"audit_status":8,"page_count":81,"language":139,"language_code":63,"site_id":62,"html_lang":63,"table_of_contents":140,"faqs":141,"seo_title":142,"seo_description":67,"update_tm":129,"read_time":39,"preview_page_count":4},962084925636,"https://ap-avatar.wpscdn.com/davatar_994ba38a5ba835b3df7d355c54d3ed8d","PSYCHO: PerSonalitY CHaracterizatiOn of artificial intelligence  \nAchal Dave Cranberry-Lemon University  \nRohit Girdhar Cranberry-Lemon University  \nAbstract  \nRecent times have seen great advancements in the field of AI, thanks to the resurgence of deep learning. It has impacted virtually every aspect of our lives, from generating new cat videos [4], to converting cat videos into dog videos [2]. However, these advancements have also stoked fear in the hearts of us humans: what if the robot hand that learned to open door knobs instead decides to use its skills to pick up a gun and point it at us? Needless to say, the solution is not fewer guns, but the mental health of these robots. In this work, we try to assuage those concerns by proposing a method to analyze the brains of our robots. Our method takes years of human psychology research and brainlessly applies it to analyze the deep networks that form the fundamental cognitive system of modern day robots. We evaluate our method on the latest and greatest deep networks and uncover the ones most likely to `break bad'.  \n1. Introduction  \n“AI is a fundamental risk to the existence of human civilization.”  \n\n| Elon Musk (July 2017)\u003Cbr>“I was trying to turn off some lights and they kept turning back on. After the third request, Alexa stopped responding and instead did an evil laugh.” |\n| --- |\n| Reddit user (January 2018)\u003Cbr>“The \\#BostonDynamics \\#robots are learning. Soon they'll be opening our fridges and stealing our beer.” |\n\nDr. Randy Olson (February 2018, via Twitter)  \nLets face it. The threat of AI is real, and the leaders of our tech industry have gone out of their way to warn us about it. However, the lack of tools to interpret our AI methods has tied the hands of AI researchers, forcing them to focus on making their methods stronger with no regard to the  \nFigure 1 . When will AI go haywire? Understanding how AI will act in the future requires a carefully designed psychological analysis using the widely acclaimed Rorschach ink blot test.  \nfuture of humanity. This problem is especially dire in the field of deep learning, where the dark magic of stochastic gradient descent carves out ultra high dimensional spaces to learn representations unimaginable by humans. In this work, we take a step back and attempt to analyze the thinking process of the deep networks we have crafted, before it is too late.  \nToday, the Turing test is largely solved [1, 5, 3] . Our method, PSYCHO instead uses the Rorschach inkblot test to analyze artificial intelligence. The test works by showing an inkblot image, like in Table 1 (column 1), and asks the user to pick a sentence that best describes that inkblot from 7 options (we follow the paradigm from [http://](http://)[ ](http://)[theinkblot.com/](theinkblot.com/)). We design an approach to allow state of the art deep networks to take this test, by finding nearest neighbors of their representation with a representation for each option. We report some insightful analysis of these networks in Sec. 3.  \n2. Approach  \nThe Rorschach ink blot test, as presented on [http:](http:)//[theinkblot.com/](theinkblot.com/), requires the test-taker to pick a sentence describing each of the 10 Rorschach ink blots. Unfortunately, despite our best efforts, we were unable to coax current AI models into taking online personality tests.  \nUndeterred, we developed a novel approach for psycho-  \nlogically evaluating our models. For each ink blot, we collected an image representing each potential response (such as “a giraffe in a bathtub”) . Unfortunately, naively collecting images can lead to a bias in the selected images. To overcome any such bias, we directly query Google Image Search for an unbiased list of images for each potential response. We then selected a single image from these results for each response query while trying our very hardest not to use our personal biases.  \nArmed with this dataset, we present each ink blot along with potential responses to our mod","cbCaiaU86Y0nPWJe","https://ap.wps.com/l/cbCaiaU86Y0nPWJe","pdf",1565698,"English","# Introduction\n## Approach\n# Experiments","[{\"question\":\"What problem does PSYCHO aim to address regarding AI advancements?\",\"answer\":\"It addresses concerns that powerful AI systems could behave in harmful ways, arguing that the key need is mental/behavioral interpretability rather than simply adding more safeguards of the same kind.\"},{\"question\":\"How does the method use the Rorschach inkblot test to analyze AI models?\",\"answer\":\"It treats each inkblot as an input and selects the best matching response option by finding nearest neighbors of the model’s representations corresponding to each candidate response.\"},{\"question\":\"Why does the approach rely on image search instead of manually collecting visuals?\",\"answer\":\"It aims to reduce bias that can arise from biased or non-representative image collection by directly querying an image search engine for images associated with each potential response.\"}]","PSYCHO - PerSonalitY CHaracterizatiOn of artificial intelligence | PDF"]