sparse attention

Een aanpak die lange inputs efficiënt verwerkt door rekenkosten te verlagen.

Articles about sparse attention

A.I
DeepSeek’s Free Models Shake Up AI Race

DeepSeek’s Free Models Shake Up AI Race
Chinese startup DeepSeek has published two new open models that claim GPT‑5‑level reasoning and long‑context abilities while cutting compute costs dramatically—a move that questio…