SubGCache reuses precomputed KV caches at the subgraph level for batches of graph RAG queries, cutting TTFT by up to 6.68x with comparable accuracy.
Batch query processing for web search engines
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
SubGCache: Accelerating Graph-based RAG with Subgraph-level KV Cache
SubGCache reuses precomputed KV caches at the subgraph level for batches of graph RAG queries, cutting TTFT by up to 6.68x with comparable accuracy.