GRINQH: Graded Input-based Quantization Hierarchy for Efficient LLM Generation Paper • 2606.23419 • Published Jun 22 • 1