LLM Routing with Dueling Feedback

• URL: http://arxiv.org/abs/2510.00841v1
• Date: Wed, 01 Oct 2025 12:52:25 GMT
• Title: LLM Routing with Dueling Feedback
• Authors: Chao-Kai Chiang, Takashi Ishida, Masashi Sugiyama,
• Abstract summary: We study the problem of selecting the best model for each query while balancing user satisfaction, model expertise, and inference cost.<n>We formulate routing as contextual dueling bandits, learning from pairwise preference feedback rather than absolute scores.<n>We introduce Category-Calibrated Fine-Tuning (CCFT), a representation-learning method that derives model embeddings from offline data using contrastive fine-tuning with categorical weighting.
• Score: 49.67815163970033
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: We study LLM routing, the problem of selecting the best model for each query while balancing user satisfaction, model expertise, and inference cost. We formulate routing as contextual dueling bandits, learning from pairwise preference feedback rather than absolute scores, thereby yielding label-efficient and dynamic adaptation. Building on this formulation, we introduce Category-Calibrated Fine-Tuning (CCFT), a representation-learning method that derives model embeddings from offline data using contrastive fine-tuning with categorical weighting. These embeddings enable the practical instantiation of Feel-Good Thompson Sampling for Contextual Dueling Bandits (FGTS.CDB), a theoretically grounded posterior-sampling algorithm. We propose four variants of the categorical weighting that explicitly integrate model quality and cost, and we empirically evaluate the proposed methods on the RouterBench and MixInstruct datasets. Across both benchmarks, our methods achieve lower cumulative regret and faster convergence, with better robustness and performance-cost balance than strong baselines built with a general-purpose OpenAI embedding model.

Related papers

• VADE: Variance-Aware Dynamic Sampling via Online Sample-Level Difficulty Estimation for Multimodal RL [38.782188833641676]
  Group-based policy optimization methods like GRPO and GSPO have become standard for training multimodal models.<n>They suffer from a critical emphgradient vanishing problem when all responses within a group receive identical rewards.<n>We propose textbfVADE, a sampling framework via online sample-level difficulty textbfEstimation.
  arXiv  Detail & Related papers  (2025-11-24T08:59:54Z)
• SPaRFT: Self-Paced Reinforcement Fine-Tuning for Large Language Models [51.74498855100541]
  Large language models (LLMs) have shown strong reasoning capabilities when fine-tuned with reinforcement learning (RL)<n>We propose textbfSPaRFT, a self-paced learning framework that enables efficient learning based on the capability of the model being trained.
  arXiv  Detail & Related papers  (2025-08-07T03:50:48Z)
• Causal LLM Routing: End-to-End Regret Minimization from Observational Data [3.3580884064577616]
  LLM routing aims to select the most appropriate model for each query.<n>Prior approaches typically adopt a decoupled strategy, where the metrics are first predicted and the model is then selected based on these estimates.<n>We propose a causal end-to-end framework that learns routing policies by minimizing decision-making regret from observational data.
  arXiv  Detail & Related papers  (2025-05-21T21:34:18Z)
• A Systematic Examination of Preference Learning through the Lens of Instruction-Following [83.71180850955679]
  We use a novel synthetic data generation pipeline to generate 48,000 instruction unique-following prompts.<n>With our synthetic prompts, we use two preference dataset curation methods - rejection sampling (RS) and Monte Carlo Tree Search (MCTS)<n>Experiments reveal that shared prefixes in preference pairs, as generated by MCTS, provide marginal but consistent improvements.<n>High-contrast preference pairs generally outperform low-contrast pairs; however, combining both often yields the best performance.
  arXiv  Detail & Related papers  (2024-12-18T15:38:39Z)
• Margin Matching Preference Optimization: Enhanced Model Alignment with Granular Feedback [64.67540769692074]
  Large language models (LLMs) fine-tuned with alignment techniques, such as reinforcement learning from human feedback, have been instrumental in developing some of the most capable AI systems to date.<n>We introduce an approach called Margin Matching Preference Optimization (MMPO), which incorporates relative quality margins into optimization, leading to improved LLM policies and reward models.<n>Experiments with both human and AI feedback data demonstrate that MMPO consistently outperforms baseline methods, often by a substantial margin, on popular benchmarks including MT-bench and RewardBench.
  arXiv  Detail & Related papers  (2024-10-04T04:56:11Z)
• Bridging and Modeling Correlations in Pairwise Data for Direct Preference Optimization [75.1240295759264]
  We propose an effective framework for Bridging and Modeling Correlations in pairwise data, named BMC.<n>We increase the consistency and informativeness of the pairwise preference signals through targeted modifications.<n>We identify that DPO alone is insufficient to model these correlations and capture nuanced variations.
  arXiv  Detail & Related papers  (2024-08-14T11:29:47Z)
• Comparative Analysis of Different Efficient Fine Tuning Methods of Large Language Models (LLMs) in Low-Resource Setting [0.0]
  We try to push the understanding of different fine-tuning strategies for large language models (LLMs) We compare state-of-the-art methods like vanilla fine-tuning and Pattern-Based Fine-Tuning (PBFT) on pre-trained models across two datasets, COLA and MNLI. Our findings suggest that these alternative strategies can exhibit out-of-domain generalization comparable to that of vanilla FT and PBFT.
  arXiv  Detail & Related papers  (2024-05-21T20:08:52Z)
• Selective Mixup Fine-Tuning for Optimizing Non-Decomposable Objectives [17.10165955576643]
  Current state-of-the-art empirical techniques offer sub-optimal performance on practical, non-decomposable performance objectives. We propose SelMix, a selective mixup-based inexpensive fine-tuning technique for pre-trained models. We find that proposed SelMix fine-tuning significantly improves the performance for various practical non-decomposable objectives across benchmarks.
  arXiv  Detail & Related papers  (2024-03-27T06:55:23Z)
• Noisy Correspondence Learning with Self-Reinforcing Errors Mitigation [63.180725016463974]
  Cross-modal retrieval relies on well-matched large-scale datasets that are laborious in practice. We introduce a novel noisy correspondence learning framework, namely textbfSelf-textbfReinforcing textbfErrors textbfMitigation (SREM)
  arXiv  Detail & Related papers  (2023-12-27T09:03:43Z)
• Optimal Sample Selection Through Uncertainty Estimation and Its Application in Deep Learning [22.410220040736235]
  We present a theoretically optimal solution for addressing both coreset selection and active learning. Our proposed method, COPS, is designed to minimize the expected loss of a model trained on subsampled data.
  arXiv  Detail & Related papers  (2023-09-05T14:06:33Z)
• A Deep Learning Method for Comparing Bayesian Hierarchical Models [1.6736940231069393]
  We propose a deep learning method for performing Bayesian model comparison on any set of hierarchical models. Our method enables efficient re-estimation of posterior model probabilities and fast performance validation prior to any real-data application.
  arXiv  Detail & Related papers  (2023-01-27T17:27:07Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.