Pinned Loading

1. perception-gap perception-gap Public

   Do frozen vision encoders trained on photos tell meaningful satellite changes from cosmetic ones? Twelve encoders, 300 EuroSAT tiles, reproducible.

   Jupyter Notebook

2. DSCadapt DSCadapt Public

   JavaScript

3. DSCadapt-backend DSCadapt-backend Public

   Python

4. ecdc-surveillance ecdc-surveillance Public

   Python

5. prior-pull prior-pull Public

   Do instruction-tuned LLMs read a person's demographics, or just react to being handed a profile? Twelve models, base and instruct paired across three labs.

   Jupyter Notebook