Dr. Adrian M. Thorne. Optimizing Large‑Scale Language Model Inference via Firmware‑Level and Architectural Attention Sparsity. Int J Mod Med 2025, 4 (10), 14-20.