publications

2025

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

   Jorge García-Carrasco, Alejandro Maté, and Juan Trujillo

   AAAI Conference on Artificial Intelligence (In Press), 2025

2024

1. How does GPT-2 Predict Acronyms? Extracting and Understanding a Circuit via Mechanistic Interpretability

   Jorge García-Carrasco, Alejandro Maté, and Juan Carlos Trujillo

   In International Conference on Artificial Intelligence and Statistics (AISTATS), 2024
2. Detecting and understanding vulnerabilities in language models via mechanistic interpretability

   Jorge García-Carrasco, Alejandro Maté, and Juan Trujillo

   In International Joint Conference on Artificial Intelligence (IJCAI), 2024