Pith. sign in

REVIEW 8 cited by

Eagle: Efficient Training-Free Router for Multi-LLM Inference

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2409.15518 v2 pith:3ITHLOYX submitted 2024-09-23 cs.LG

Eagle: Efficient Training-Free Router for Multi-LLM Inference

classification cs.LG
keywords eaglemodelonlinebaselineefficientenvironmentshigh-volumemethods
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

The proliferation of Large Language Models (LLMs) with varying capabilities and costs has created a need for efficient model selection in AI systems. LLM routers address this need by dynamically choosing the most suitable model for a given query based on task requirements and budget constraints. However, existing routers face challenges in scalability and real-time adaptation, particularly in high-volume online environments. We present Eagle, a novel LLM routing approach that combines global and local ELO ranking modules to overcome these limitations. By evaluating both general and specialized LLM abilities, Eagle provides a scalable, training-free solution that enhances model selection quality while reducing computational overhead. Our experiments across multiple datasets show Eagle consistently outperforms baseline methods, with improvements of up to 23.52 percent in Area Under Curve (AUC) scores. Moreover, Eagle demonstrates remarkable efficiency, requiring only 1/20 of baseline methods' time for initialization and 100 to 200 times faster incremental updates in online scenarios, making it well-suited for dynamic, high-volume online serving environments.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 8 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Continual Model Routing in Evolving Model Hubs

    cs.AI 2026-05 unverdicted novelty 7.0

    Formalizes continual model routing (CMR), releases CMRBench with over 2000 models, and presents CARvE which outperforms retrieval, fine-tuning and adapter-merging baselines on model/family/domain accuracy.

  2. Route to Rome Attack: Directing LLM Routers to Expensive Models via Adversarial Suffix Optimization

    cs.CR 2026-04 unverdicted novelty 7.0

    R²A uses a hybrid ensemble surrogate router and suffix optimization to significantly increase black-box LLM router selection of expensive models across query distributions.

  3. Rethinking Predictive Modeling for LLM Routing: When Simple kNN Beats Complex Learned Routers

    cs.LG 2025-05 conditional novelty 7.0

    A well-tuned kNN router matches or exceeds state-of-the-art learned routers on new standardized benchmarks spanning instruction, QA, reasoning, and the first multi-modal visual routing dataset, due to locality of mode...

  4. The Routing Plateau: Understanding and Breaking the Accuracy Limits of LLM Routers

    cs.LG 2026-05 unverdicted novelty 6.0

    LLM routers across 21 methods on 5 benchmarks converge to similar accuracy below oracle due to learning global performance trends rather than fine-grained query signals.

  5. SynConfRoute: Syntax-Aware Routing for Efficient Code Completion with Small CodeLLMs

    cs.SE 2026-05 unverdicted novelty 6.0

    SynConfRoute routes code completions using syntax validation and token confidence, improving pass@1 by up to 31% on hard tasks and reducing accelerator usage by 58% versus always using the largest model.

  6. From Sampled Outcomes to Capability Distributions: Rethinking Supervision for LLM Routing

    cs.LG 2026-06 unverdicted novelty 5.0

    DARS replaces single-shot response labels with distribution-aware supervision derived from input and output uncertainty to produce more reliable LLM routing policies.

  7. Training-Free versus Training-Based Intent Classification in LLMs: Accuracy, Robustness, and Failure Modes

    cs.CL 2026-08 conditional novelty 4.0

    Statistical classifiers built on LLM activation norms and coordinates match or beat trained MLP heads on coarse intent routing and resist camouflage better, while MLPs win on fine-grained subfield distinctions.

  8. Harnessing Multiple Large Language Models: A Survey on LLM Ensemble

    cs.CL 2025-02 unverdicted novelty 2.0

    A systematic survey of LLM ensemble methods organized into a taxonomy of ensemble-before-inference, ensemble-during-inference, and ensemble-after-inference stages, with review of benchmarks, applications, and future d...