Extracting Interpretable Task-Specific Circuits from Large Language Models for Faster Inference.

Jorge Garca-Carrasco, Alejandro Mat, Juan Trujillo

Browse the full AAAI paper archive.