Der Workshop richtet sich an Promovierende in jeder Phase ihrer Promotion sowie an Masterabsolvent*innen, die eine Promotion in Erwägung ziehen. Er bietet die Möglichkeit, eigene Forschungsprojekte oder -vorhaben vorzustellen und in einer unterstützenden, kollaborativen Atmosphäre zu diskutieren.\n \nPräsentationen:\nAnastasia Glawatzki (Universität Münster): \"Wissensumgebung und Wissenspraktiken im Zeitalter des AI Slops\"\nAnna-Maria Linstaedt (Heinrich-Heine-Universität Düsseldorf): \"Digitales Community Management von staatlichen Akteuren in sozialen Medien\"\nJana Peters (Heinrich-Heine-Universität Düsseldorf): \"Künstliche Intelligenz als Akteur deliberativer Öffentlichkeit: Psychologische Mechanismen und empirische Wirkungen KI-gestützter Interventionen in digitalen Diskursen\"

As part of the annual convention of the DGPuK Division \"Digital Communication,\" we invite you to join our Workshop on \"Large Language Models & Users\' Mental Models.\" LLMs are transforming how individuals seek andprocess information, construct their environment and themselves, and evaluate truth in public discourse. Our focus is on users\' mental models of LLMs, regarding epistemology (truth, trust, knowledge), design (algorithms, interfaces, AI systems), and effects (cognition, emotion, behavior). We bring together researchers from communication science, psychology, computer science/HCI, and philosophy, and welcome perspectives from beyond these fields.\nMental models increasingly form through interaction with digital environments rather than direct experience, a process social media accelerated by circulating simplified, emotionally salient representations. LLMs further transform this landscape by generatingcoherent, context-sensitive representations on demand, shaping the construction, reinforcement, and revision of users\' mental models. This workshop zooms in on these developments, exploring themes such as human–AI interaction and mental models, epistemic trust and credibility in AI systems, disinformation, narratives, and bias, algorithmic design and explainability, domain-specific communication (health and nutrition, politics, sustainability), and emotional and behavioral effects. \n\nThe workshop language is English; the venue is wheelchair accessible. The workshop is open to all conference participants. Please register at eres@tu-berlin.de.\n\nProgram overview: https://docs.google.com/document/d/1o2jx8cyn6NtzlnGf_evE-ZU9r6GzeRCK/edit?usp=sharing&ouid=110708348742398302550&rtpof=true&sd=true

This room can be used throughout the conference as a quiet space to take a short break or work. It is equipped with a large round table, several chairs, and electrical outlets.