4 papers
Concept-Level Explainability for Auditing & Steering LLM Responses
Kenza Amara, Rita Sevastjanova, Mennatallah El-Assady
As large language models (LLMs) become widely deployed, concerns about their safety and alignment grow. An approach to steer LLM behavior, such as mitigating biases or defending ag…
ReforesTree: A Dataset for Estimating Tropical Forest Carbon Stock with Deep Learning and Aerial Imagery
Gyri Reiersen, David Dao, Björn Lütjens +5
Forest biomass is a key influence for future climate, and the world urgently needs highly scalable financing schemes, such as carbon offsetting certifications, to protect and resto…
PowerGraph: A power grid benchmark dataset for graph neural networks
Anna Varbella, Kenza Amara, Blazhe Gjorgiev +2
Power grids are critical infrastructures of paramount importance to modern society and, therefore, engineered to operate under diverse conditions and failures. The ongoing energy t…
Why context matters in VQA and Reasoning: Semantic interventions for VLM input modalities
Kenza Amara, Lukas Klein, Carsten Lüth +3
The various limitations of Generative AI, such as hallucinations and model failures, have made it crucial to understand the role of different modalities in Visual Language Model (V…