BEGIN:VCALENDAR VERSION:2.0 PRODID:IEEE vTools.Events//EN CALSCALE:GREGORIAN BEGIN:VTIMEZONE TZID:Europe/Vienna BEGIN:DAYLIGHT DTSTART:20260329T030000 TZOFFSETFROM:+0100 TZOFFSETTO:+0200 RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=3 TZNAME:CEST END:DAYLIGHT BEGIN:STANDARD DTSTART:20251026T020000 TZOFFSETFROM:+0200 TZOFFSETTO:+0100 RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10 TZNAME:CET END:STANDARD END:VTIMEZONE BEGIN:VEVENT DTSTAMP:20260313T175323Z UID:1276B292-1E1F-4C82-89D0-4DECC4FEC2CD DTSTART;TZID=Europe/Vienna:20260310T170000 DTEND;TZID=Europe/Vienna:20260310T183000 DESCRIPTION:Dear colleague\,\n\nWe are pleased to invite you to this talk i n our Lecture Series:\n\nTuesday\, March 10\, 2026 at 5:00 p.m. (17:00) Ce ntral European Time (UTC+1)\n\nTopic: “LLMs When Left Alone”\n(scroll down for abstract and CV)\n\nSpeaker: Stefan Szeider (TU Wien\, AT)\nModer ator: Julia Neidhardt (TU Wien\, AT)\n\nTo participate in the talks via Zo om go to: [https://tuwien.zoom.us/j/96389928143?pwd=UU5YRkNuRmdoWHV4MFBwMW RCcUErdz09](https://eur01.safelinks.protection.outlook.com/?url=http%3A%2F %2Fxuxss.mjt.lu%2Flnk%2FAU8AAI_pMJcAAc5b9QQAAFNIx9MAAAAARcEAn-g1ABG7CQBpoI MEbK4FgGI0Q0mVQAn6DiiGfQARx9M%2F1%2Fw70eJxCrM8AuiVBmyi5SxA%2FaHR0cHM6Ly90d XdpZW4uem9vbS51cy9qLzk2Mzg5OTI4MTQzP3B3ZD1VVTVZUmtOdVJtZG9XSFY0TUZCd01XUkN jVUVyZHowOQ&data=05%7C02%7Cmichael.heiss%40siemens.com%7C1ddbcb00f88d435f6 e7708de755c9fbf%7C38ae3bcd95794fd4addab42e1495d55a%7C1%7C0%7C6390772378416 86175%7CUnknown%7CTWFpbGZsb3d8eyJFbXB0eU1hcGkiOnRydWUsIlYiOiIwLjAuMDAwMCIs IlAiOiJXaW4zMiIsIkFOIjoiTWFpbCIsIldUIjoyfQ%3D%3D%7C0%7C%7C%7C&sdata=%2BGp8 hBsSPhI%2BMKPPzunDOWfdCJr7Lsxp6WVOCyMor4U%3D&reserved=0)\n(Password: 0dzqx qiy)\n\nThe talk will be live streamed and recorded on our YouTube channel : [https://www.youtube.com/digitalhumanism](https://eur01.safelinks.protec tion.outlook.com/?url=http%3A%2F%2Fxuxss.mjt.lu%2Flnk%2FAU8AAI_pMJcAAc5b9Q QAAFNIx9MAAAAARcEAn-g1ABG7CQBpoIMEbK4FgGI0Q0mVQAn6DiiGfQARx9M%2F2%2FbAkwQ9 TOOQ43eGKUaGHynA%2FaHR0cHM6Ly93d3cueW91dHViZS5jb20vZGlnaXRhbGh1bWFuaXNt&da ta=05%7C02%7Cmichael.heiss%40siemens.com%7C1ddbcb00f88d435f6e7708de755c9fb f%7C38ae3bcd95794fd4addab42e1495d55a%7C1%7C0%7C639077237841731249%7CUnknow n%7CTWFpbGZsb3d8eyJFbXB0eU1hcGkiOnRydWUsIlYiOiIwLjAuMDAwMCIsIlAiOiJXaW4zMi IsIkFOIjoiTWFpbCIsIldUIjoyfQ%3D%3D%7C0%7C%7C%7C&sdata=K9HU0cFNomqv5l0vpza9 8VSkU8dQovkcdzl%2FTiig3o8%3D&reserved=0)\n\nFor further announcements and information about the speakers in the Lecture Series\, see [https://dighum .org/#latest-news](https://eur01.safelinks.protection.outlook.com/?url=htt p%3A%2F%2Fxuxss.mjt.lu%2Flnk%2FAU8AAI_pMJcAAc5b9QQAAFNIx9MAAAAARcEAn-g1ABG 7CQBpoIMEbK4FgGI0Q0mVQAn6DiiGfQARx9M%2F3%2FcD8nRML9lUw78mcuBRHTRg%2FaHR0cH M6Ly9kaWdodW0ub3JnLyNsYXRlc3QtbmV3cw%23latest-news&data=05%7C02%7Cmichael. heiss%40siemens.com%7C1ddbcb00f88d435f6e7708de755c9fbf%7C38ae3bcd95794fd4a ddab42e1495d55a%7C1%7C0%7C639077237841768458%7CUnknown%7CTWFpbGZsb3d8eyJFb XB0eU1hcGkiOnRydWUsIlYiOiIwLjAuMDAwMCIsIlAiOiJXaW4zMiIsIkFOIjoiTWFpbCIsIld UIjoyfQ%3D%3D%7C0%7C%7C%7C&sdata=51%2BEXIH9VafI8Tt3BAnzw%2Fl6b5fEUk6ODSir0 gNKYoU%3D&reserved=0). Please note that you can access the slides and reco rdings of our past events via that link.\n\nIn case you missed the last le cture by Sayeed Choudhury you can watch the recording of [“The Imperativ e of Openness in AI”](https://eur01.safelinks.protection.outlook.com/?ur l=http%3A%2F%2Fxuxss.mjt.lu%2Flnk%2FAU8AAI_pMJcAAc5b9QQAAFNIx9MAAAAARcEAn- g1ABG7CQBpoIMEbK4FgGI0Q0mVQAn6DiiGfQARx9M%2F4%2FuCh4FMibdryY3dJzTw2dNg%2Fa HR0cHM6Ly95b3V0dS5iZS9xVk5CWUtlenNZMA&data=05%7C02%7Cmichael.heiss%40sieme ns.com%7C1ddbcb00f88d435f6e7708de755c9fbf%7C38ae3bcd95794fd4addab42e1495d5 5a%7C1%7C0%7C639077237841809359%7CUnknown%7CTWFpbGZsb3d8eyJFbXB0eU1hcGkiOn RydWUsIlYiOiIwLjAuMDAwMCIsIlAiOiJXaW4zMiIsIkFOIjoiTWFpbCIsIldUIjoyfQ%3D%3D %7C0%7C%7C%7C&sdata=AilVwUELvkZhwOy%2B8kGlk%2FxjYklXLvGIsvLg7go%2B%2Fmw%3D &reserved=0).\n\nAnnouncement of our next event:\n21.04.26 DigHum Lecture Darja Djordjevic on AI companionship\nreminder our new public calendar: [h ttps://calendar.google.com/calendar/ical/dighumtuwien%40gmail.com/public/b asic.ics](https://eur01.safelinks.protection.outlook.com/?url=http%3A%2F%2 Fxuxss.mjt.lu%2Flnk%2FAU8AAI_pMJcAAc5b9QQAAFNIx9MAAAAARcEAn-g1ABG7CQBpoIME bK4FgGI0Q0mVQAn6DiiGfQARx9M%2F5%2FAUFXZDIEk0y2E-ww8IjjRw%2FaHR0cHM6Ly9jYWx lbmRhci5nb29nbGUuY29tL2NhbGVuZGFyL2ljYWwvZGlnaHVtdHV3aWVuJTQwZ21haWwuY29tL 3B1YmxpYy9iYXNpYy5pY3M&data=05%7C02%7Cmichael.heiss%40siemens.com%7C1ddbcb 00f88d435f6e7708de755c9fbf%7C38ae3bcd95794fd4addab42e1495d55a%7C1%7C0%7C63 9077237841849162%7CUnknown%7CTWFpbGZsb3d8eyJFbXB0eU1hcGkiOnRydWUsIlYiOiIwL jAuMDAwMCIsIlAiOiJXaW4zMiIsIkFOIjoiTWFpbCIsIldUIjoyfQ%3D%3D%7C0%7C%7C%7C&s data=oXpGGhWMLz68lXWoI3kzuZbjTX8DWIiQsnbAJfk55Hw%3D&reserved=0)\n\nWe are looking forward to seeing you!\n\nCo-sponsored by: DIGHUM\n\nAgenda: \nABS TRACT “LLMs When Left Alone”:\n\nToday\, autonomous AI agents are so w idely deployed that millions of people interact with them daily. These inc lude coding assistants\, chatbots\, and systems that run for hours without human oversight. A natural question to ask is whether something is going on inside these systems beyond sophisticated pattern matching. The questio n is easy to dismiss but hard to settle. The standard dismissals (anthropo morphism\, reductionism\, training bias) each carry force\, but none is co nclusive. In this talk\, we follow an empirical approach. Instead of specu lating\, we placed frontier language models in minimal autonomous loops wi th no task and no user. We observe that the behavioral patterns that emerg ed are surprisingly stable and model-specific. The patterns proved stable enough to identify the source model with high accuracy. We further examine d whether AI self-reports of internal states can be trusted. We find that a “placebo tool” that does nothing reliably shifts what models say abo ut themselves. These findings raise questions that go beyond computer scie nce and concern society at large. These questions will only grow more pres sing as AI agents become more autonomous.\n\nShort Bio of Stefan Szeider:\ n\nStefan Szeider is an Austrian computer scientist and full professor at the Faculty of Informatics at the Vienna University of Technology (TU Wien )\, where he chairs the Algorithms and Complexity Group. He is also a visi ting scientist at the Simons Institute for the Theory of Computing at UC B erkeley. Specializing in algorithms\, logic\, computational complexity\, a nd theoretical computer science\, Szeider designs efficient algorithms for challenging problems in Artificial Intelligence and Logic. With a doctora te from the University of Vienna\, he has held positions at universities i n the UK and Canada. Szeider co-founded the Vienna Center for Logic and Al gorithms\, was the first Austrian computer scientist to receive an ERC Sta rting Grant\, and has published over 250 articles in scientific journals a nd conference proceedings.\n\nShort Bio of Julia Neidhardt:\n\nJulia Neidh ardt is an Assistant Professor at the Research Unit Data Science at TU Wie n Informatics with a background in mathematics and computer science. Her w ork focuses on user modeling and recommender systems in e-commerce\, touri sm and news\, online opinion-forming and behavior\, and Digital Humanism. She has published in leading venues\, including Nature Human Behaviour\, a nd is regularly invited to speak internationally. She is a senior program committee member of ACM RecSys\, associate editor of the Journal of Inform ation Technology & Tourism\, and a distinguished reviewer for ACM TORS. Sh e leads the Christian Doppler Laboratory for Recommender Systems at TU Wie n and\, since 2023\, serves as UNESCO Co-Chair on Digital Humanism.\n\nVir tual: https://events.vtools.ieee.org/m/542812 LOCATION:Virtual: https://events.vtools.ieee.org/m/542812 ORGANIZER:m.heiss@ieee.org SEQUENCE:8 SUMMARY:IEEE TEMS/LMAG: "LLMs When Left Alone” (DigHum-Series) URL;VALUE=URI:https://events.vtools.ieee.org/m/542812 X-ALT-DESC:Description: <br /><p>Dear colleague\,</p>\n<p>We are pleased to invite you to this talk in our Lecture Series:</p>\n<p><strong>Tuesday\, March 10\, 2026 at 5:00 p.m.</strong> (17:00) Central European Time (UTC+1 )</p>\n<p>Topic: <strong>&ldquo\;LLMs When Left Alone&rdquo\;</strong><br> (scroll down for abstract and CV)</p>\n<p>Speaker: <strong>Stefan Szeider< /strong> (TU Wien\, AT)<br>Moderator: <strong>Julia Neidhardt</strong> (TU Wien\, AT)</p>\n<p>To participate in the talks via <strong>Zoom</strong> go to: <a href="https://eur01.safelinks.protection.outlook.com/?url=http%3 A%2F%2Fxuxss.mjt.lu%2Flnk%2FAU8AAI_pMJcAAc5b9QQAAFNIx9MAAAAARcEAn-g1ABG7CQ BpoIMEbK4FgGI0Q0mVQAn6DiiGfQARx9M%2F1%2Fw70eJxCrM8AuiVBmyi5SxA%2FaHR0cHM6L y90dXdpZW4uem9vbS51cy9qLzk2Mzg5OTI4MTQzP3B3ZD1VVTVZUmtOdVJtZG9XSFY0TUZCd01 XUkNjVUVyZHowOQ&amp\;data=05%7C02%7Cmichael.heiss%40siemens.com%7C1ddbcb00 f88d435f6e7708de755c9fbf%7C38ae3bcd95794fd4addab42e1495d55a%7C1%7C0%7C6390 77237841686175%7CUnknown%7CTWFpbGZsb3d8eyJFbXB0eU1hcGkiOnRydWUsIlYiOiIwLjA uMDAwMCIsIlAiOiJXaW4zMiIsIkFOIjoiTWFpbCIsIldUIjoyfQ%3D%3D%7C0%7C%7C%7C&amp \;sdata=%2BGp8hBsSPhI%2BMKPPzunDOWfdCJr7Lsxp6WVOCyMor4U%3D&amp\;reserved=0 ">https://tuwien.zoom.us/j/96389928143?pwd=UU5YRkNuRmdoWHV4MFBwMWRCcUErdz0 9</a> <br>(Password: 0dzqxqiy)</p>\n<p>The talk will be live streamed and recorded on our <strong>YouTube</strong> channel: <a href="https://eur01.s afelinks.protection.outlook.com/?url=http%3A%2F%2Fxuxss.mjt.lu%2Flnk%2FAU8 AAI_pMJcAAc5b9QQAAFNIx9MAAAAARcEAn-g1ABG7CQBpoIMEbK4FgGI0Q0mVQAn6DiiGfQARx 9M%2F2%2FbAkwQ9TOOQ43eGKUaGHynA%2FaHR0cHM6Ly93d3cueW91dHViZS5jb20vZGlnaXRh bGh1bWFuaXNt&amp\;data=05%7C02%7Cmichael.heiss%40siemens.com%7C1ddbcb00f88 d435f6e7708de755c9fbf%7C38ae3bcd95794fd4addab42e1495d55a%7C1%7C0%7C6390772 37841731249%7CUnknown%7CTWFpbGZsb3d8eyJFbXB0eU1hcGkiOnRydWUsIlYiOiIwLjAuMD AwMCIsIlAiOiJXaW4zMiIsIkFOIjoiTWFpbCIsIldUIjoyfQ%3D%3D%7C0%7C%7C%7C&amp\;s data=K9HU0cFNomqv5l0vpza98VSkU8dQovkcdzl%2FTiig3o8%3D&amp\;reserved=0">htt ps://www.youtube.com/digitalhumanism</a></p>\n<p style="margin-bottom: 12. 0pt\;">For further announcements and information about the speakers in the Lecture Series\, see <a href="https://eur01.safelinks.protection.outlook. com/?url=http%3A%2F%2Fxuxss.mjt.lu%2Flnk%2FAU8AAI_pMJcAAc5b9QQAAFNIx9MAAAA ARcEAn-g1ABG7CQBpoIMEbK4FgGI0Q0mVQAn6DiiGfQARx9M%2F3%2FcD8nRML9lUw78mcuBRH TRg%2FaHR0cHM6Ly9kaWdodW0ub3JnLyNsYXRlc3QtbmV3cw%23latest-news&amp\;data=0 5%7C02%7Cmichael.heiss%40siemens.com%7C1ddbcb00f88d435f6e7708de755c9fbf%7C 38ae3bcd95794fd4addab42e1495d55a%7C1%7C0%7C639077237841768458%7CUnknown%7C TWFpbGZsb3d8eyJFbXB0eU1hcGkiOnRydWUsIlYiOiIwLjAuMDAwMCIsIlAiOiJXaW4zMiIsIk FOIjoiTWFpbCIsIldUIjoyfQ%3D%3D%7C0%7C%7C%7C&amp\;sdata=51%2BEXIH9VafI8Tt3B Anzw%2Fl6b5fEUk6ODSir0gNKYoU%3D&amp\;reserved=0">https://dighum.org/#lates t-news</a>. Please note that you can access the slides and recordings of o ur past events via that link.<br><br>In case you missed the last lecture b y Sayeed Choudhury you can watch the recording of <a href="https://eur01.s afelinks.protection.outlook.com/?url=http%3A%2F%2Fxuxss.mjt.lu%2Flnk%2FAU8 AAI_pMJcAAc5b9QQAAFNIx9MAAAAARcEAn-g1ABG7CQBpoIMEbK4FgGI0Q0mVQAn6DiiGfQARx 9M%2F4%2FuCh4FMibdryY3dJzTw2dNg%2FaHR0cHM6Ly95b3V0dS5iZS9xVk5CWUtlenNZMA&a mp\;data=05%7C02%7Cmichael.heiss%40siemens.com%7C1ddbcb00f88d435f6e7708de7 55c9fbf%7C38ae3bcd95794fd4addab42e1495d55a%7C1%7C0%7C639077237841809359%7C Unknown%7CTWFpbGZsb3d8eyJFbXB0eU1hcGkiOnRydWUsIlYiOiIwLjAuMDAwMCIsIlAiOiJX aW4zMiIsIkFOIjoiTWFpbCIsIldUIjoyfQ%3D%3D%7C0%7C%7C%7C&amp\;sdata=AilVwUELv kZhwOy%2B8kGlk%2FxjYklXLvGIsvLg7go%2B%2Fmw%3D&amp\;reserved=0">&ldquo\;The Imperative of Openness in AI&rdquo\;</a>.<br><br><br>Announcement of our next event:<br>21.04.26 DigHum Lecture Darja Djordjevic on AI companionshi p <br>reminder our new public calendar: <a href="https://eur01.safelinks.p rotection.outlook.com/?url=http%3A%2F%2Fxuxss.mjt.lu%2Flnk%2FAU8AAI_pMJcAA c5b9QQAAFNIx9MAAAAARcEAn-g1ABG7CQBpoIMEbK4FgGI0Q0mVQAn6DiiGfQARx9M%2F5%2FA UFXZDIEk0y2E-ww8IjjRw%2FaHR0cHM6Ly9jYWxlbmRhci5nb29nbGUuY29tL2NhbGVuZGFyL2 ljYWwvZGlnaHVtdHV3aWVuJTQwZ21haWwuY29tL3B1YmxpYy9iYXNpYy5pY3M&amp\;data=05 %7C02%7Cmichael.heiss%40siemens.com%7C1ddbcb00f88d435f6e7708de755c9fbf%7C3 8ae3bcd95794fd4addab42e1495d55a%7C1%7C0%7C639077237841849162%7CUnknown%7CT WFpbGZsb3d8eyJFbXB0eU1hcGkiOnRydWUsIlYiOiIwLjAuMDAwMCIsIlAiOiJXaW4zMiIsIkF OIjoiTWFpbCIsIldUIjoyfQ%3D%3D%7C0%7C%7C%7C&amp\;sdata=oXpGGhWMLz68lXWoI3kz uZbjTX8DWIiQsnbAJfk55Hw%3D&amp\;reserved=0">https://calendar.google.com/ca lendar/ical/dighumtuwien%40gmail.com/public/basic.ics</a><br><br>We are lo oking forward to seeing you!</p><br /><br />Agenda: <br /><p><strong>ABSTR ACT &ldquo\;LLMs When Left Alone&rdquo\;:</strong></p>\n<p>Today\, autonom ous AI agents are so widely deployed that millions of people interact with them daily. These include coding assistants\, chatbots\, and systems that run for hours without human oversight. A natural question to ask is wheth er something is going on inside these systems beyond sophisticated pattern matching. The question is easy to dismiss but hard to settle. The standar d dismissals (anthropomorphism\, reductionism\, training bias) each carry force\, but none is conclusive. In this talk\, we follow an empirical appr oach. Instead of speculating\, we placed frontier language models in minim al autonomous loops with no task and no user. We observe that the behavior al patterns that emerged are surprisingly stable and model-specific. The p atterns proved stable enough to identify the source model with high accura cy. We further examined whether AI self-reports of internal states can be trusted. We find that a &ldquo\;placebo tool&rdquo\; that does nothing rel iably shifts what models say about themselves. These findings raise questi ons that go beyond computer science and concern society at large. These qu estions will only grow more pressing as AI agents become more autonomous.< /p>\n<p><strong>Short Bio of Stefan Szeider:</strong></p>\n<p>Stefan Szeid er is an Austrian computer scientist and full professor at the Faculty of Informatics at the Vienna University of Technology (TU Wien)\, where he ch airs the Algorithms and Complexity Group. He is also a visiting scientist at the Simons Institute for the Theory of Computing at UC Berkeley. Specia lizing in algorithms\, logic\, computational complexity\, and theoretical computer science\, Szeider designs efficient algorithms for challenging pr oblems in Artificial Intelligence and Logic. With a doctorate from the Uni versity of Vienna\, he has held positions at universities in the UK and Ca nada. Szeider co-founded the Vienna Center for Logic and Algorithms\, was the first Austrian computer scientist to receive an ERC Starting Grant\, a nd has published over 250 articles in scientific journals and conference p roceedings.</p>\n<p><strong>Short Bio of Julia Neidhardt:</strong></p>\n<p >Julia Neidhardt is an Assistant Professor at the Research Unit Data Scien ce at TU Wien Informatics with a background in mathematics and computer sc ience. Her work focuses on user modeling and recommender systems in e-comm erce\, tourism and news\, online opinion-forming and behavior\, and Digita l Humanism. She has published in leading venues\, including Nature Human B ehaviour\, and is regularly invited to speak internationally. She is a sen ior program committee member of ACM RecSys\, associate editor of the Journ al of Information Technology &amp\; Tourism\, and a distinguished reviewer for ACM TORS. She leads the Christian Doppler Laboratory for Recommender Systems at TU Wien and\, since 2023\, serves as UNESCO Co-Chair on Digital Humanism.</p> END:VEVENT END:VCALENDAR