发表机构
Zhejiang University; Hangzhou High-Tech Zone (Binjiang) Institute of Blockchain and Data Security（浙江大学; 杭州高新区（滨江）区块链与数据安全研究院）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
提出QCATS框架，利用查询上下文感知的Transformer切片实现数据库内高效稀疏推理，通过查询级路由和系统优化，在保持精度的同时降低延迟达4.42倍。
AI 中文摘要
数据库内预测性查询处理越来越多地在关系型流水线中应用基于Transformer的模型。然而，现有的数据库内推理通常仅向推理运行时暴露元组级别的模型输入，使得关系谓词和元数据统计信息对神经执行规划不可见。在本文中，我们提出了QCATS，一种查询上下文感知的Transformer切片框架，能够在数据库系统内部实现高效的稀疏推理。QCATS在查询粒度上执行：在推理过程中，它不是路由单个令牌或元组，而是利用查询谓词和元数据统计信息，在模型执行前预先选择与上下文对齐的FFN切片。该框架包含离线专家构建和轻量级查询级路由，后者在执行期间动态选择专家。QCATS进一步引入了系统优化，包括异步CPU-GPU流水线和路由感知的批处理。在四个预测性查询工作负载上使用BERT-base和Qwen-0.6B进行的实验表明，QCATS在保持与密集基线相当的预测准确性的同时，实现了高达4.42倍的延迟降低。
英文摘要
In-database predictive query processing increasingly applies Transformer-based models within relational pipelines. However, existing in-database inference typically exposes only tuple-level model inputs to the inference runtime, leaving relational predicates and metadata statistics invisible to neural execution planning. In this paper, we propose QCATS, a query context-aware transformer slicing framework that enables efficient sparse inference inside database systems. QCATS executes at query granularity: instead of routing individual tokens or tuples during inference, it uses query predicates and metadata statistics to pre-select context-aligned FFN slices before model execution. The framework comprises offline expert construction and lightweight query-level routing that dynamically selects experts during execution. QCATS further introduces system optimizations, including asynchronous CPU-GPU pipelines and routing-aware batching. Experiments on four predictive-query workloads with BERT-base and Qwen-0.6B show that QCATS achieves up to 4.42x latency reduction while preserving prediction accuracy comparable to dense baselines.
