paper-with-me

Papers

LLM Routing as Reasoning: A MaxSAT View

2026-03-13 · Son Nguyen, Xinyuan Liu, Ransalu Senanayake arxiv

Routing a query through an appropriate LLM is challenging, particularly when user preferences are expressed in natural language and model attributes are only partially observable. We propose a constraint-based interpretation of language-conditioned LLM routing, formulating it as a weighted MaxSAT/MaxSMT problem in which natural language feedback induces hard and soft constraints over model attributes. Under this view, routing corresponds to selecting models that approximately maximize satisfaction of feedback-conditioned clauses. Empirical analysis on a 25-model benchmark shows that language feedback produces near-feasible recommendation sets, while no-feedback scenarios reveal systematic priors. Our results suggest that LLM routing can be understood as structured constraint optimization under language-conditioned preferences.

📄 PDF Abstract BibTeX arXiv:2603.13612

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Certifying Pareto-Optimality in Multi-Objective Maximum Satisfiability

2025-01-29 · Christoph Jabs, Jeremias Berg, Bart Bogaerts, Matti Järvisalo

Due to the wide employment of automated reasoning in the analysis and construction of correct systems, the results reported by automated reasoning engines must be trustworthy. For Boolean satisfiability (SAT) solvers - a…

DPMS: An ADD-Based Symbolic Approach for Generalized MaxSAT Solving

2022-05-08 · Anastasios Kyrillidis, Moshe Y. Vardi, Zhiwei Zhang

Boolean MaxSAT, as well as generalized formulations such as Min-MaxSAT and Max-hybrid-SAT, are fundamental optimization problems in Boolean reasoning. Existing methods for MaxSAT have been successful in solving benchmark…

Reliable Reasoning with Large Language Models via Preference-Based Maximum Satisfiability

2026-05-28 · Pedro Orvalho, Marta Kwiatkowska, Guillem Alenyà, Felip Manyà arxiv

Large Language Models (LLMs) excel at understanding natural language but struggle with optimisation tasks involving multiple constraints and user-defined preferences, which commonly arise in domains such as robotics. We …

Code Generation

MaxSAT-Based Feedback for Guiding Vision-Language Models in Sudoku

2026-07-14 · Pedro Orvalho, Guillem Alenyà, Felip Manyà arxiv

Vision--Language Models (VLMs) have recently demonstrated promising performance on structured visual reasoning tasks, including grid-based puzzles. However, despite strong perceptual capabilities, these models lack expli…

Visual Reasoning

Can Graph Neural Networks Learn to Solve MaxSAT Problem?

2021-11-15 · Minghao Liu, Fuqi Jia, Pei Huang, Fan Zhang 외

With the rapid development of deep learning techniques, various recent work has tried to apply graph neural networks (GNNs) to solve NP-hard problems such as Boolean Satisfiability (SAT), which shows the potential in bri…