Sparsity-Guided Holistic Explanation for LLMs with Interpretable Inference-Time Intervention

Open in new window