dynamic-sparsification

Tag

Cards List
#dynamic-sparsification

Self-Pruned Key-Value Attention: Learning When to Write by Predicting Future Utility

arXiv cs.LG · 2026-05-15 Cached

Introduces Self-Pruned Key-Value Attention (SP-KV), a mechanism that learns to predict future utility of key-value pairs to dynamically prune the KV cache, reducing memory usage and decoding speed by 3-10x with minimal performance degradation. The model and utility predictor are trained end-to-end using next-token prediction.

0 favorites 0 likes
#dynamic-sparsification

UniPrefill: Universal Long-Context Prefill Acceleration via Block-wise Dynamic Sparsification

arXiv cs.CL · 2026-05-08 Cached

UniPrefill is a new prefill acceleration framework proposed in a research paper that enables block-wise dynamic sparsification for universal long-context processing in LLMs. It integrates with vLLM to achieve up to 2.1x speedup in Time-To-First-Token across various model architectures.

0 favorites 0 likes
← Back to home

Submit Feedback