Beyond Static RAG: An Adaptive, Tri-Metric Routing Framework for Efficient Long-Context Inference on Commodity GPUs

Abstract:Deploying retrieval-augmented generation (RAG) on commodity GPUs such as the NVIDIA T4 (16 GB VRAM) exposes a practical failure mode we call the Compression Paradox: neural prompt compression can add key-value (KV) cache contention and preprocessing latency that outweigh generation-time savings, while skipping compression can cause out-of-memory (OOM) failures on long contexts. We identify two distinct failure mechanisms when a vLLM-served LLM and a PyTorch-based compressor are co-deployed under tight memory budgets, and introduce the Tri-Metric Router, a deterministic, training-free policy that selects among Raw, Neural (LLMLingua-2), and Lexical (BM25) pipelines. The router uses three CPU-side signals: spatial complexity ($L$), syntactic density ($\rho_{key}$), and type-token ratio (TTR). Unlike prior semantic-only adaptation, our dispatch signal is hardware-physical, based on VRAM headroom and a latency crossover point. Thresholds are calibrated from profiling on LongBench qasper, yielding an operating crossover near 4,332 words on T4; our contribution is this calibration methodology rather than a hardware-specific constant. On out-of-distribution holdouts, the method achieves 0% OOM failures, 88.5 $\pm$ 4.4% oracle alignment, and 49.3% Combined F1, improving over always-on lexical compression by 5.2 points without additional VRAM or training cost.

This Paper is accepted and presented at ICML Scale Workshop 2026. this https URL (Paper ID :72)

Machine Learning (cs.LG); Information Retrieval (cs.IR)

arXiv:2609.17564 [cs.LG]

(or arXiv:2609.17564v1 [cs.LG] for this version)

https://doi.org/10.48550/arXiv.2609.17564 Focus to learn more arXiv-issued DOI via DataCite

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.