Two Papers Accepted in NeurIPS 2024
Two papers on “PrivCirNet: Efficient Private Inference via Block Circulant Transformation” and “ArkVale: Efficient Generative LLM Inference with Recallable Key-Value Eviction” are accpted by NeurIPs'2024.
Two papers on “PrivCirNet: Efficient Private Inference via Block Circulant Transformation” and “ArkVale: Efficient Generative LLM Inference with Recallable Key-Value Eviction” are accpted by NeurIPs'2024.