MPCache: MPC-Friendly KV Cache Eviction for Efficient Private Large Language Model Inference

Re-assign community