# query

Model Extraction from Multimodal Systems

Techniques for extracting model capabilities, weights, and architecture details from multimodal AI systems through visual, audio, and cross-modal query strategies.