JanusQuant: Accurate and Efficient 2-bit KV Cache Quantization for Long-Context Inference

Long-context large language models (LLMs) have seen widespread adoption in recent years. However, during inference, the key-value (KV) cache—which stores intermediate activations—consumes significant memory, particularly as sequence lengths grow. Quantization offers a promising path to compress KV cache, but existing 2-bit approaches fall short of achieving optimal inference efficiency due to hardware-unfriendly algorithms and system implementations.