LT-TTD 的理论保证：一种用于两级排序系统的统一 Transformer 架构
Theoretical Guarantees for LT-TTD: A Unified Transformer-based Architecture for Two-Level Ranking Systems
AI总结：
本文提出统一两级排序的 LT-TTD 架构，结合双塔模型与 Transformer，并给出误差传播、排序质量和收敛性的理论保证，同时引入 UPQE 评估指标。
摘要：
现代推荐与搜索系统通常采用多阶段排序架构，以高效处理数十亿候选对象。传统方法使用具有不同优化目标的独立 L1（候选检索）和 L2（重排序）模型，这带来了不可逆误差传播和次优排序等关键局限。本文识别并分析了这种解耦范式的根本局限，并提出 LT-TTD（Listwise Transformer with Two-Tower Distillation，带双塔蒸馏的列表级 Transformer），一种连接检索与排序阶段的创新统一架构。我们的方法在统一列表级学习框架中结合了双塔模型的计算效率和 Transformer 的表达能力。我们对架构进行了全面的理论分析，并在误差传播抑制、排序质量提升和优化收敛方面建立了形式化保证。我们推导出的理论界限表明，LT-TTD 将不可检索相关项目的上限降低了一个取决于知识蒸馏强度的因子，并证明我们的多目标优化框架相比分离训练能够达到可证明的更优全局最优。此外，我们分析了方法的计算复杂度，证明其渐近复杂度在实际应用中保持在可行范围内。我们还引入了 UPQE，一种专为统一排序架构设计的新型评估指标，能够整体性地刻画检索质量、排序性能和计算效率。
英文摘要：
Modern recommendation and search systems typically employ multi-stage ranking architectures to efficiently handle billions of candidates. The conventional approach uses distinct L1 (candidate retrieval) and L2 (re-ranking) models with different optimization objectives, introducing critical limitations including irreversible error propagation and suboptimal ranking. This paper identifies and analyzes the fundamental limitations of this decoupled paradigm and proposes LT-TTD (Listwise Transformer with Two-Tower Distillation), a novel unified architecture that bridges retrieval and ranking phases. Our approach combines the computational efficiency of two-tower models with the expressivity of transformers in a unified listwise learning framework. We provide a comprehensive theoretical analysis of our architecture and establish formal guarantees regarding error propagation mitigation, ranking quality improvements, and optimization convergence. We derive theoretical bounds showing that LT-TTD reduces the upper limit on irretrievable relevant items by a factor that depends on the knowledge distillation strength, and prove that our multi-objective optimization framework achieves a provably better global optimum than disjoint training. Additionally, we analyze the computational complexity of our approach, demonstrating that the asymptotic complexity remains within practical bounds for real-world applications. We also introduce UPQE, a novel evaluation metric specifically designed for unified ranking architectures that holistically captures retrieval quality, ranking performance, and computational efficiency.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
