EconBase
← All papers

Online Pandora's Box for Contextual LLM Cascading

Alexandre Belloni, Yan Chen, Yehua Wei

arXiv 5 Jun 2026 · Artificial Intelligence

arXiv:2606.07392 · PDF · DOI · OpenAlex · Extracted main text

Abstract

Motivated by Large Language Model (LLM) cascading, we propose an online contextual Pandora's Box model for adaptively querying and selecting LLM APIs. In each period, a decision-maker observes a request context and faces a two-phase decision problem. In the query phase, the decision-maker sequentially queries APIs, where each query reveals a generated output and the decision-maker incurs an (output-dependent) cost. In the selection phase, the decision-maker selects one of the generated outputs to deploy and observes only the downstream reward of the deployed output. This output-mediated feedback structure differs from classical online contextual Pandora's Box models, in which opening a box directly reveals its reward. Rather than estimating the full conditional output and cost distributions of each API, we directly model the reservation index and develop a learning approach for the query phase. Specifically, we impose a parametric structure on the contextual reservation index functions induced by the classical Weitzman's policy. Our policy combines generalized method of moments (GMM) type estimation of these reservation indices with UCB-style confidence bounds for both these indices and the shared output-level reward evaluator. Under regularity conditions, we prove that the resulting policy achieves dimension-dependent $\widetilde O(\sqrt T)$ cumulative regret over a horizon of $T$ periods.

Citation extraction

72
references
107
in-text mentions
72
distinct cited
1
self-citations
31,253
main-text words

appendix boundary found by none_found · 100% of the source is main text. Read the extracted text to check this.

Most heavily cited references

The works this paper leans on most, across its whole bibliography — not restricted to papers in our corpus. Ranked by composite intensity, which combines how often a work is mentioned, how many sections mention it, and how much of that falls in the main text rather than the appendix.

ReferenceIntensityMentionsSectionsMain text
1Chen, Lingjiao and Zaharia, Matei and Zou, James (2025) FrugalGPT: How to Use Large Language Models While Reducing Cost and Improving Performance1.00073100%
2Filippi, Sarah and Cappe, Olivier and Garivier, Aurélien and Szepesv… (2010) Parametric bandits: The generalized linear case1.00073100%
3Abbasi-Yadkori, Yasin and Pál, Dávid and Szepesvári, Csaba (2011) Improved algorithms for linear stochastic bandits0.92843100%
4Weitzman, Martin L (1979) OPTIMAL SEARCH FOR THE BEST ALTERNATIVE.0.92843100%
5Atsidakou, Alexia and Caramanis, Constantine and Gergatsouli, Evange… (2024) Contextual pandora’s box0.87482100%
6Gupta, Neha and Narasimhan, Harikrishna and Jitkrittum, Wittawat and… (2024) Language model cascades: Token-level uncertainty and beyond0.58531100%
7Yue, Murong and Zhao, Jie and Zhang, Min and Du, Liang and Yao, Ziyu (2024) Large Language Model Cascades with Mixture of Thought Representations for Cost-Efficient Reasoning0.58531100%
8Freedman, David A (1975) On tail probabilities for martingales0.51121100%
9Gergatsouli, Evangelia and Tzamos, Christos (2022) Online learning for min sum set cover and pandora’s box0.51121100%
10Lee, Junghyun and Yun, Se-Young and Jun, Kwang-Sung (2024) A Unified Confidence Sequence for Generalized Linear Models, with Applications to Bandits0.51121100%

Showing the top 10 of 72 scored citations.