ReasonCache reuses similar KV cache states across reasoning steps in LRMs via collaborative filtering to boost serving throughput by up to 89.2% while preserving accuracy.
H.; Gonzalez, J.; Zhang, H.; and Stoica, I
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
ReasonCache: Accelerating Large Reasoning Model Serving through KV Cache Sharing
ReasonCache reuses similar KV cache states across reasoning steps in LRMs via collaborative filtering to boost serving throughput by up to 89.2% while preserving accuracy.