Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Are Optimal Algorithms Still Optimal? Rethinking Sorting in LLM-Based Pairwise Ranking with Batching and Caching

About

We introduce a novel framework for analyzing sorting algorithms in pairwise ranking prompting (PRP), re-centering the cost model around LLM inferences rather than traditional pairwise comparisons. While classical metrics based on comparison counts have traditionally been used to gauge efficiency, our analysis reveals that expensive LLM inferences overturn these predictions; accordingly, our framework encourages strategies such as batching and caching to mitigate inference costs. We show that algorithms optimal in the classical setting can lose efficiency when LLM inferences dominate the cost under certain optimizations.

Juan Wisznia, Cecilia Bola\~nos, Juan Tollo, Giovanni Marraffini, Agust\'in Gianolini, Noe Hsueh, Luciano Del Corro• 2025

Related benchmarks

TaskDatasetResultRank
RerankingTREC DL 2020
NDCG@100.7045
132
RerankingTREC DL 2019 (test)
NDCG@1067.12
108
Document RerankingTREC DL 2019 and 2020 (test)
NDCG@1065.42
108
RerankingTREC DL 2019 v1 (test)
NDCG@1065.73
108
Document RerankingTREC DL 19
NDCG@1067.44
39
RerankingBEIR (test)
Covid Score74.8
19
End-to-end RerankingTREC DL 2019 2020 Average
Average NDCG@1068.95
10
Showing 7 of 7 rows

Other info

Follow for update