INFERENCEDYNAMICS: Efficient Routing Across LLMs through Structured Capability and Knowledge Profiling

TL;DR

InferenceDynamics is a scalable routing framework that models LLM capabilities to efficiently select optimal models for diverse tasks, improving performance and resource use across large pools of models.

Contribution

The paper introduces a novel multi-dimensional routing framework that models model capabilities and knowledge, addressing scalability and adaptability issues in LLM routing.

Findings

Effective group-level routing demonstrated on multiple benchmarks.

Superior task performance with optimized model selection.

Framework generalizes well across diverse LLMs and tasks.

Abstract

Large Language Model (LLM) routing is a pivotal technique for navigating a diverse landscape of LLMs, aiming to select the best-performing LLMs tailored to the domains of user queries, while managing computational resources. However, current routing approaches often face limitations in scalability when dealing with a large pool of specialized LLMs, or in their adaptability to extending model scope and evolving capability domains. To overcome those challenges, we propose InferenceDynamics, a flexible and scalable multi-dimensional routing framework by modeling the capability and knowledge of models. We operate it on our comprehensive dataset RouteMix, and demonstrate its effectiveness and generalizability in group-level routing using modern benchmarks including MMLU-Pro, GPQA, BigGenBench, and LiveBench, showcasing its ability to identify and leverage top-performing models for given tasks, leading to superior outcomes with efficient resource utilization. The broader adoption of Inference Dynamics can empower users to harness the full specialized potential of the LLM ecosystem, and our code will be made publicly available to encourage further research.