AI.info

Xiaofei He

Explore Xiaofei He on AI.info.

• HeteroCache: A Dynamic Retrieval Approach to Heterogeneous KV Cache Compression for Long-Context LLM Inference