Online Learning with LLM Experts from Limited Feedback
Not on the map. Why?
Abstract (arXiv)
We study adaptive routing of prompts to large language model (LLM) experts to maximize response quality in an online setting with limited feedback. We formulate it as a bandit problem with K actions that represent experts and d features that encode prompts, over a horizon of T rounds. We propose algorithms that strategically select and observe rewards to minimize regret. In the full-information setting, we achieve a regret of tildeO(d T / √m), while in the bandit setting we achieve tildeO(d T √K / m), where m ll T is a budget on feedback. Our experiments show that we efficiently learn high-quality routing strategies across diverse LLMs from limited feedback.
Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Builds on 22 of its 50 references that are in the map. First posted on arXiv 5 Sep 2026.
Builds on
22 of 50 references.
References · 22
- 2023FrugalGPT: How to Use Large Language Models While Reducing Cost and Improving Performancecited by 853
- 2024RouteLLM: Learning to Route LLMs with Preference Datacited by 682
- 2024From Generation to Judgment: Opportunities and Challenges of LLM-as-a-judgecited by 494
- 2024LLMs-as-Judges: A Comprehensive Survey on LLM-based Evaluation Methodscited by 465
- 2024Hybrid LLM: Cost-Efficient and Quality-Aware Query Routingcited by 372
- 2023Large Language Model Routing with Benchmark Datasetscited by 172
- 2023Fly-Swat or Cannon? Cost-Effective Language Model Choice via Meta-Modelingcited by 95
- 2015Linear Contextual Bandits with Knapsackscited by 82
- 2015Cascading Bandits: Learning to Rank in the Cascade Modelcited by 77
- 2019Model selection for contextual banditscited by 60
- 2016Conservative Banditscited by 59
- 2016Conservative Contextual Linear Banditscited by 57
10 more
- 2018Linear Bandits with Stochastic Delayed Feedbackcited by 55
- 2024MetaLLM: A High-performant and Cost-efficient Dynamic Framework for Wrapping LLMscited by 41
- 2024Cost-Effective Online Multi-LLM Selection with Versatile Reward Modelscited by 39
- 2024Towards Optimizing the Costs of LLM Usagecited by 33
- 2024Optimising Calls to Large Language Models with Uncertainty-Based Two-Tier Selectioncited by 33
- 2017Online Learning with Abstentioncited by 28
- 2024Which LLM to Play? Convergence-Aware Online Model Selection with Time-Increasing Banditscited by 27
- 2023Large Language Model-Enhanced Algorithm Selection: Towards Comprehensive Algorithm Representationcited by 21
- 2020Online Active Model Selection for Pre-trained Classifierscited by 20
- 2025ThriftLLM: On Cost-Effective Selection of Large Language Models for Classification Queriescited by 14
Built on by
Citing papers per year: 2026: 1.
1 paper.