1 paper · 1 filter
Yirui Liu, Ruoling Qi, Longwen Wang +5
LLM serving is increasingly accelerated by position-independent caching (PIC). Existing PIC methods, however, are built for full-attention models, where a token-indexed KV cache un…