I am a Postdoctoral Researcher at the Center for AI in Society (CAISA), University of Copenhagen. I am affiliated with the LAMP group, CoAStaL NLP, and the Pioneer Centre for AI. My research focuses on developing and evaluating vision-and-language models, with a particular interest in multimodal understanding and reasoning, AI safety, robustness, and responsible AI.
I completed my Ph.D. in Natural Language Processing at the University of Sheffield (UK), supervised by Professor Nikos Aletras, where I worked on computational social science, political communication, and multimodal learning. During my Ph.D., I was also an Applied Scientist Intern at Amazon Alexa Shopping.
Research interests: Multimodal AI, natural language processing, trustworthy AI, multimodal understanding and reasoning, evaluation, robustness, and AI safety.
Community: I am an active member of LatinX in AI and an organizer of SomosNLP, where I support researchers, encourage collaboration, and help build a more connected Ibero-American NLP community.
Email: davi@samf.ku.dk
[Blog] [Publications] [Google Scholar]
News
- Aug 2026
- Started a new position as a Postdoctoral Researcher at CAISA, University of Copenhagen. Excited to continue my research on vision-language models with a focus on trustworthiness and safety.
- Our new pre-print is available: What’s the Catch? Evaluating Temporal Consistency in Vision-Language Models
- Jul 2026
- Our paper Reasoning Dynamics and the Limits of Monitoring Modality Reliance in Vision-Language Models has been accepted to appear at COLM 2026 🥳 See you in San Francisco!
- 🎉 Our work, Beyond Binary Classification: Detecting Fine-Grained Sexism in Social Media Videos, has been published in Computational Linguistics.