Cassia Sampaio Sanctos

Cassia Sampaio Sanctos

Auto-generated profile, based on publicly available information

Cassia Sampaio Sanctos

São Paulo, IBM Research, São Paulo, Brazil

Research Focus & Metrics

Academic Career Length

4Years (2023-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
25 April 2025
Verified
Conference Paper
Open access
VSVagner Figueredo de Santana
Vagner Figueredo de Santana
SBSara E Berger
Sara E Berger
HCHeloisa Candello
Heloisa Candello
TMTiago Machado
Tiago Machado
CSCassia Sampaio Sanctos
Cassia Sampaio Sanctos
LWLemara Williams
Lemara Williams

24 March 2025
Verified
Conference Paper
Open access
VSVagner Figueredo de Santana
Vagner Figueredo de Santana
SBSara Berger
Sara Berger
TMTiago Machado
Tiago Machado
MMMaysa Malfiza Garcia de Macedo
Maysa Malfiza Garcia de Macedo
CSCassia Sampaio Sanctos
Cassia Sampaio Sanctos
ZWZhaoqing Wu
Zhaoqing Wu

19 July 2023
Verified
Conference Paper
Open access
HCHeloisa Candello
Heloisa Candello
GSGabriel Meneguelli Soella
Gabriel Meneguelli Soella
CSCassia Sampaio Sanctos
Cassia Sampaio Sanctos
MGMarcelo Carpinette Grave
Marcelo Carpinette Grave
AFAdinan Alves De Brito Filho
Adinan Alves De Brito Filho

8 September 2025
Verified
Conference Paper
Closed access
VSVagner Figueredo de Santana
Vagner Figueredo de Santana
AAAshwath Vaithinathan Aravindan
Ashwath Vaithinathan Aravindan
CSCassia Sampaio Sanctos
Cassia Sampaio Sanctos
TMTiago Machado
Tiago Machado
LSLuan Soares de Souza
Luan Soares de Souza
Introdução: Large Language Models (LLMs) estão sendo propostos como solução para múltiplos fluxos de trabalho, mas eles comumente não contam com orientação adequada a usuários e nem fornecem informações sobre IA Responsável no momento da criação de prompts. Objetivo: Nesse contexto, esta pesquisa propõe uma forma de orientar e conscientizar pessoas sobre IA Responsável enquanto interagem com LLMs em múltiplos turnos. Metodologia ou Etapas: Em nossa ferramenta, usuários recebem sugestões de IA Responsável enquanto escrevem prompts para LLMs. Resultados: Espera-se que este trabalho motive mais sistemas de recomendação de prompting para promover IA Responsável no momento da criação dos prompts.

29 July 2026 in SSRN Electronic Journal
Verified
Preprint
Open access
VSVagner Figueredo de Santana
Vagner Figueredo de Santana
CSCassia Sampaio Sanctos
Cassia Sampaio Sanctos
AAAshwath Vaithinathan Aravindan
Ashwath Vaithinathan Aravindan
Implementing Responsible Artificial Intelligence (RAI) requires addressing incentives,responsibility gaps between technology creation and use, power dynamics, and criticalreflection throughout projects. To understand how technology creators critically reflecton different forces in place (self, business, and society) when contributing to an open-source project. Understanding how technology creators embed values into code isessential for supporting RAI. This study is grounded on the R&I Formative Frameworkand employs Computational Ground Theory to assess 194 engagements (i.e., comments,questions, and issues) with an open source RAI project after people become awareabout the project when attending talks/tutorials on the said project. Findings suggestthat contributors envisioning new contexts of use and indirect impacts of technology(e.g., diverse languages, accessibility needs, potential harms, varied domains, or businessapplications) were more likely to critically reflect and embed their values while askingquestions, ideating, or coding. In contrast, discussions limited to technical aspects didnot reach RAI considerations in depth. Hence, we argue that tools, coding practices,collaborative methods, and open source communities should foster technology creatorsto share their views and experiences and consciously think about (in)direct impacts oftechnology and more diverse contexts of creation and use.