QV-PIC, a rendered-image RAG caching framework, achieves higher F1 and lower time-to-first-token than text-PIC baselines by combining chat-template-conditioned cache compilation with query-aware dual-resolution selection.
InAdvances in Neural Information Processing Systems, volume 33, 9459–9474
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
QV-PIC: Query-Aware Visual Position-Independent Caching for Efficient RAG Serving
QV-PIC, a rendered-image RAG caching framework, achieves higher F1 and lower time-to-first-token than text-PIC baselines by combining chat-template-conditioned cache compilation with query-aware dual-resolution selection.