prefix-inference

Tag

Cards List
#prefix-inference

Context Memorization for Efficient Long Context Generation

Hugging Face Daily Papers · 2026-05-18 Cached

Proposes attention-state memory, a training-free approach that stores precomputed attention states in lightweight memory to improve accuracy and reduce latency for long prefix inference, outperforming traditional methods on benchmarks.

0 favorites 0 likes
← Back to home

Submit Feedback