We won! 🥳 Award for best scientific paper at an international conference with over 1,000 entries!

I am deeply honored to receive the 2026 CSCWD/IEEE Best Conference Paper Award! ❤️🏆

Presentation and Acceptance of the Best Conference Paper Award

Our research on cognitive biases in Large Language Models (LLMs) found that confirmation bias was present in all evaluated models, ranging from 15% to 67%;. The causes may stem from the dataset, the Transformer architecture, the configuration, or the prompt engineering itself, and that “assumption”-type interactions lead to twice as many errors as “doubt”-type interactions. Particularly in the healthcare context, where this technology integrates Clinical Decision Support Systems (CDSS), the validation process becomes essential to ensure patient safety and the reliability of AI-supported decision-making.

This is a tremendous achievement for me and for the following institutions: UFRJ (Federal University of Rio de Janeiro), NCE / UFRJ Tércio Pacitti Institute, INESC TEC, and the University of Trás-os-Montes and Alto Douro! ❤️ I would like to express my deepest gratitude to my advisors, Hugo Paredes and Adriana Vivacqua, for their encouragement and for the lessons they have taught me both in the classroom and in life! ❤️ To the other co-authors Gonçalo Gonçalves, Leonardo Lopes, Tiago Dantas, Dennis Paulino, A. Thiago Netto, Diogo Guimarães, OLAVO TEIXEIRA, and Artur Rocha, for all your support. ❤️ And for the exchange of experiences over the years with professors from whom I have learned so much: Jano Moreira de Souza, Weiming Shen, Marcos Kalinowski, Angelica Dias, Juliana França, Ana Cristina Bicharra Garcia, Claudio Miceli de Farias, Tânia Rocha, and Geraldo Xexéo.

The article is titled “Confirmation Bias in Large Language Models: Risks for Clinical Decision Support Systems” and will be published soon!

Written by Prof. Rafael Ris-Ala, translated from:

https://academicool.blogspot.com/2026/05/vencemos-premio-de-melhor-artigo.html