跳到正文
arXiv cs.AI· Benjamin Clavi\'e, Herumb Shandilya, Hanna Lichtenberg, Aamir Shakir, Makoto P. Kato·· 2 小时前AI 评分42

DealRank:改进 Listwise LLM 重排调度的新方法

Deal With It: On Better Schedules for Listwise Reranking

AI 导读

一篇 arXiv 论文证明,在并行调度下的 Listwise LLM 重排中,按检索器排名轮发(round-robin)分配候选是最优切分方式,并据此提出 DealRank 调度。

正文 · 原文

View PDF HTML (experimental)

Abstract:Listwise LLM rerankers cannot reliably rank hundreds of documents in a single call, and rely on schedules to orchestrate their calls. Parallel schedules split candidates into groups and advance a fixed quota from each. When a group holds more of the top passages than its quota, some are cut, and expensive repair calls become required. We show that the way candidates are split matters where such a quota can bind, and that, in expected cuts, the best split there deals them round-robin by retriever rank, like cards. Although not conceptually new, dealing has remained unanalyzed. We prove that, under a simple retriever model, no split with equal quotas cuts fewer of the true top ten on average, and that contiguous chunks cut more as the retriever improves. DealRank, the simplest schedule our analysis allows, is on average within 0.1 nDCG@10 points of the best schedules over six LLMs, in two rounds. It is consistently 1.1 to 1.4 times faster than BlitzRank with concurrent first calls up to a thousand candidates. There, it needs three rounds and is eleven times faster than the sliding window. In existing schedules such as BlitzRank, dealing reduces repair calls from 59% of queries to 14%, improving latency with no significant change in quality. Random partitioning appears equivalent at a quota of four, as predicted, but ties results to the seed and trails by 1.3 nDCG@10 points at a quota of one over a dense retriever.
Subjects: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI)
Cite as: arXiv:2611.22392 [cs.IR]
  (or arXiv:2611.22392v1 [cs.IR] for this version)
  https://doi.org/10.48550/arXiv.2611.22392

arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Benjamin Clavié [view email]
[v1] Tue, 17 Nov 2026 08:42:11 UTC (3,016 KB)

来源:arXiv cs.AI · arxiv.org