DuoAttention identifies retrieval heads requiring full KV cache and streaming heads using constant-length cache to reduce memory and latency in long-context LLM inference.
Fast transformer decoding: One write-head is all you need, 2019
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
roles
background 1polarities
background 1representative citing papers
Putri is a structured pruning technique for LLMs that compensates for pruning errors via weight updates and sequential processing while pruning at the attention-head level to reach state-of-the-art results at extreme sparsity.
citing papers explorer
-
DuoAttention: Efficient Long-Context LLM Inference with Retrieval and Streaming Heads
DuoAttention identifies retrieval heads requiring full KV cache and streaming heads using constant-length cache to reduce memory and latency in long-context LLM inference.
-
Prune, Update and Trim: Robust Structured Pruning for Large Language Models
Putri is a structured pruning technique for LLMs that compensates for pruning errors via weight updates and sequential processing while pruning at the attention-head level to reach state-of-the-art results at extreme sparsity.
- The General Theory of Localization Methods