EXIST

This lab focuses on the detection of sexist messages in social networks (SN). Inequality and discrimination against women that remains embedded in society is increasingly being replicated online. Internet perpetuates and even naturalizes gender differences and sexist attitudes. The EXIST 2026 lab will continue to focus on the detection of sexism in social networks, while introducing a novel paradigm that integrates human-centered signals into the AI development pipeline. In this edition, we extend the “learning with disagreement” framework by incorporating sensor-based data from people exposed to potentially sexist content. This includes measurements such as skin conductance, heart rate variability, and other sensor data that reflect unconscious responses to sexism. Given the nature of these multimodal signals, we will concentrate on analyzing memes and short videos—formats that combine visual and textual cues and are especially suited for capturing the emotional and cognitive impact of online content. This human-in-the-loop approach not only acknowledges the diversity of subjective reactions to sexism, but also opens new avenues for building more robust, equitable, and interpretable systems. By integrating both conscious feedback and unconscious reactions from annotators, EXIST 2026 aims to foster a more nuanced and ethically grounded understanding of sexism across platforms and formats.

Organizers

• Jorge Carrillo-De-Albornoz (Universidad Nacional de Educacion a Distancia)
• Laura Plaza (Universidad Nacional de Educación a Distancia)
• Damiano Spina (RMIT University)
• Paolo Rosso (Universitat Politècnica de València)
• Iván Arcos (Universitat Politècnica de València)
• Elena Gomis (Universitat Politècnica de València)
• María Aloy (Universitat Politècnica de València)