← 研踪 PaperTrailarXiv:2611.23898 · v1
Cascadeability and the Realization Gap: An Empirical Study of LLM-Powered Semantic Filtering
首次提交 2026-11-18 13:50
版本更新 2026-11-18 13:50
雷达观测 2026-11-19 19:08
摘要
Semantic filters enable natural language predicates over unstructured documents, but evaluating every document with a large language model (LLM) oracle is expensive. Proxy cascades reduce oracle invocations using cheaper surrogates, but oracle reduction alone cannot distinguish proxy limitations from calibration bottlenecks and establish practical efficiency. We present a unified benchmark of four representative systems, which formulate query processing as a three-stage workflow comprising proxy construction, threshold calibration, and cascade inference, using a two-threshold acceptance rule. We introduce cascadeability: the maximum empirical oracle reduction attainable from fixed proxy scores at a specified accuracy target, computed retrospectively using complete oracle labels. We investigate 8 research questions, spanning proxy quality, end-to-end performance, workload sensitivities, and component ablations, and reveal our key findings accordingly. Comparing cascadeability with achieved oracle reduction yields a realization gap that, interpreted alongside accuracy, exposes distinct proxy and calibration bottlenecks. The benchmark and diagnostics support evaluation beyond oracle reduction and motivate joint proxy-calibration design, class-sensitive quality assessment, and end-to-end cost optimization. We release a dataset-query benchmark, containing 10 real-world workloads across various domains and difficulties, with annotations of two LLM families.
作者备注:8 pages. Code and data available at https://github.com/Anto-an/semantic_filter_benchmark
观测到的版本
相邻记录可能跳过中间版本;首次发现 v2 及以上也会标为版本更新。
v1 · 2026-11-18 13:50 · 最新观测版本
Cascadeability and the Realization Gap: An Empirical Study of LLM-Powered Semantic Filtering
Semantic filters enable natural language predicates over unstructured documents, but evaluating every document with a large language model (LLM) oracle is expensive. Proxy cascades reduce oracle invocations using cheaper surrogates, but oracle reduction alone cannot distinguish proxy limitations from calibration bottlenecks and establish practical efficiency. We present a unified benchmark of four representative systems, which formulate query processing as a three-stage workflow comprising proxy construction, threshold calibration, and cascade inference, using a two-threshold acceptance rule. We introduce cascadeability: the maximum empirical oracle reduction attainable from fixed proxy scores at a specified accuracy target, computed retrospectively using complete oracle labels. We investigate 8 research questions, spanning proxy quality, end-to-end performance, workload sensitivities, and component ablations, and reveal our key findings accordingly. Comparing cascadeability with achieved oracle reduction yields a realization gap that, interpreted alongside accuracy, exposes distinct proxy and calibration bottlenecks. The benchmark and diagnostics support evaluation beyond oracle reduction and motivate joint proxy-calibration design, class-sensitive quality assessment, and end-to-end cost optimization. We release a dataset-query benchmark, containing 10 real-world workloads across various domains and difficulties, with annotations of two LLM families.
查看此版本解读 →查看此版本原文 ↗