AI.info

Jian Jiang

Explore Jian Jiang on AI.info.

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