7 papers
Evaluating the Interpretability of Sparse Autoencoders with Concept Annotations
Jonas Klotz, Cassio F. Dantas, Pallavi Jain +2
Sparse autoencoders (SAEs) are increasingly used to extract interpretable concepts from vision and vision language models, yet existing evaluation methods largely rely on proxy met…
RAMEN: Resolution-Adjustable Multimodal Encoder for Earth Observation
Nicolas Houdré, Diego Marcos, Hugo Riffaud de Turckheim +4
Earth observation (EO) data spans a wide range of spatial, spectral, and temporal resolutions, from high-resolution optical imagery to low resolution multispectral products or rada…
Metonymy in vision models undermines attention-based interpretability
Ananthu Aniraj, Cassio F. Dantas, Dino Ienco +2
Part-based reasoning is a classical strategy to make a computer vision model directly focus on the object parts that are relevant to the downstream task. In the context of deep lea…
Two-stage Vision Transformers and Hard Masking offer Robust Object Representations
Ananthu Aniraj, Cassio F. Dantas, Dino Ienco +1
Context can strongly affect object representations, sometimes leading to undesired biases, particularly when objects appear in out-of-distribution backgrounds at inference. At the…
TimeSenCLIP: A Time Series Vision-Language Model for Remote Sensing
Pallavi Jain, Diego Marcos, Dino Ienco +2
Vision-language models (VLMs) have shown significant promise in remote sensing applications, particularly for land-use and land-cover (LULC) mapping via zero-shot classification an…
Atomizer: Generalizing to new modalities by breaking satellite images down to a set of scalars
Hugo Riffaud de Turckheim, Sylvain Lobry, Roberto Interdonato +1
The growing number of Earth observation satellites has led to increasingly diverse remote sensing data, with varying spatial, spectral, and temporal configurations. Most existing m…