Liyuan Liu

2

Papers

382

Total Citations

Papers (2)

Model Tells You What to Discard: Adaptive KV Cache Compression for LLMs

Vector-ICL: In-context Learning with Continuous Vector Representations