6 papers
Risk Governance for Generative AI Mental Health Support: A Multi-Turn Safety Architecture
Anabela C. Areias, Catarina Botelho, António Farinhas +7
Large language models (LLMs) are increasingly used for emotional support despite lacking mechanisms to safely govern evolving mental health risk. Existing safety approaches primari…
MindGuard: Guardrail Classifiers for Multi-Turn Mental Health Support
António Farinhas, Nuno M. Guerreiro, José Pombal +6
Large language models are increasingly used for mental health support, yet their conversational coherence alone does not ensure clinical appropriateness. Existing general-purpose s…
MindEval: Benchmarking Language Models on Multi-turn Mental Health Support
José Pombal, Maya D'Eon, Nuno M. Guerreiro +3
Demand for mental health support through AI chatbots is surging, though current systems present several limitations, like sycophancy or overvalidation, and reinforcement of maladap…
Movie Facts and Fibs (MF): A Benchmark for Long Movie Understanding
Emmanouil Zaranis, António Farinhas, Saul Santos +28
Despite recent progress in vision-language models (VLMs), holistic understanding of long-form video content remains a significant challenge, partly due to limitations in current be…
-Video: A Training-Free Approach to Long Video Understanding via Continuous-Time Memory Consolidation
Saul Santos, António Farinhas, Daniel C. McNamee +1
Current video-language models struggle with long-video understanding due to limited context lengths and reliance on sparse frame subsampling, often leading to information loss. Thi…
Translate Smart, not Hard: Cascaded Translation Systems with Quality-Aware Deferral
António Farinhas, Nuno M. Guerreiro, Sweta Agrawal +2
Larger models often outperform smaller ones but come with high computational costs. Cascading offers a potential solution. By default, it uses smaller models and defers only some i…