Pith. sign in

REVIEW 16 cited by

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 2502.08773 v2 pith:4JIKGC4N submitted 2025-02-12 cs.CL cs.LG

Universal Model Routing for Efficient LLM Inference

classification cs.CL cs.LG
keywords routingllmsunirouteinferencemodelpoolproblemamongst
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
read the original abstract

Model routing is a simple technique for reducing the inference cost of large language models (LLMs), wherein one maintains a pool of candidate LLMs, and learns to route each prompt to the smallest feasible LLM. Existing works focus on learning a router for a fixed pool of LLMs. In this paper, we consider the problem of dynamic routing, where new, previously unobserved LLMs are available at test time. We propose UniRoute, a new approach to this problem that relies on representing each LLM as a feature vector, derived based on predictions on a set of representative prompts. Based on this, we detail two effective instantiations of UniRoute, relying on cluster-based routing and a learned cluster map respectively. We show that these are estimates of a theoretically optimal routing rule, and quantify their errors via an excess risk bound. Experiments on a range of public benchmarks show the effectiveness of UniRoute in routing amongst more than 30 unseen LLMs.

discussion (0)

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

Forward citations

Cited by 16 Pith papers

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

  1. Rosetta Memory: Adaptive Memory for Cross-LLM Agents

    cs.LG 2026-06 unverdicted novelty 7.0

    Rosetta Memory trains two profile-conditioned operators with a minimum-gain sampling curriculum and performance-gap reward to enable memory transfer between LLMs, showing gains on multi-hop QA benchmarks and robustnes...

  2. Latency-Quality Routing for Functionally Equivalent Tools in LLM Agents

    cs.LG 2026-05 unverdicted novelty 7.0

    LQM-ContextRoute routes LLM tool calls via latency-quality matching in a contextual bandit, improving F1 by 2.18 pp, accuracy by up to 18 pp, and NDCG by 2.91-3.22 pp over SW-UCB on web-search, StrategyQA, and retriev...

  3. Flexible Routing via Uncertainty Decomposition

    cs.LG 2026-05 unverdicted novelty 7.0

    A router that decomposes uncertainty to flexibly route queries between cheap models and oracles while providing regret bounds and supporting abstention in classification tasks with multiple annotations.

  4. Switchcraft: AI Model Router for Agentic Tool Calling

    cs.AI 2026-05 unverdicted novelty 7.0

    Switchcraft routes agentic tool-calling queries to the lowest-cost model that preserves correctness, reaching 82.9% accuracy and 84% cost reduction on five benchmarks.

  5. SecureRouter: Encrypted Routing for Efficient Secure Inference

    cs.CR 2026-04 unverdicted novelty 7.0

    SecureRouter accelerates secure transformer inference by 1.95x via an encrypted router that selects input-adaptive models from an MPC-optimized pool with negligible accuracy loss.

  6. Prompting Complexity: Shortest Prompts for Texts and Behaviors in LLMs

    cs.CL 2026-07 conditional novelty 6.0

    The paper defines prompting complexity as the length of the shortest plausible prompt that deterministically generates a target text with a fixed language model.

  7. SWE-Router: Routing in Multi-turn Agentic Software Engineering Tasks

    cs.SE 2026-06 unverdicted novelty 6.0

    SWE-Router introduces trajectory-conditioned value-based routing for LLM agents on SWE tasks, with a Bayes-optimality theorem and empirical cost savings while retaining most strong-model performance.

  8. The Capability Frontier: Benchmarks Miss 82% of Model Performance

    cs.AI 2026-06 unverdicted novelty 6.0

    The Capability Frontier shows standard LLM benchmarks underestimate performance by 82% by ignoring model specialization across questions and the benefits of multiple generations per query.

  9. RouteBalance: Fused Model Routing and Load Balancing for Heterogeneous LLM Serving

    cs.DC 2026-06 unverdicted novelty 6.0

    RouteBalance fuses routing and load balancing for heterogeneous LLM serving and traces the upper quality-cost-throughput frontier on a 13-instance 28-GPU cluster.

  10. 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.

  11. Latency-Quality Routing for Functionally Equivalent Tools in LLM Agents

    cs.LG 2026-05 unverdicted novelty 6.0

    LQM-ContextRoute routes tool calls by expected quality per service cycle using contextual bandits and LLM-as-judge feedback, yielding +2.18 pp F1, up to +18 pp accuracy, and +2.91-3.22 pp NDCG gains over SW-UCB on web...

  12. Learning Agent Routing From Early Experience

    cs.CL 2026-05 unverdicted novelty 6.0

    BoundaryRouter routes queries to LLM or agent using early experience memory from a seed set, cutting inference time 60.6% versus always using agents and raising performance 28.6% versus always using direct LLM inference.

  13. Cost-Ordered Feasibility for Multi-Armed Bandits with Cost Subsidy

    cs.LG 2026-05 unverdicted novelty 6.0

    Develops COF algorithm for MAB-CS that intelligently checks cheap arm feasibility by pooling samples, with generalized instance-dependent lower bounds and matching upper bounds on cumulative cost and quality regret.

  14. Privacy-Preserving LLMs Routing

    cs.CR 2026-04 unverdicted novelty 6.0

    PPRoute achieves plaintext-level LLM routing quality with MPC-based privacy and a 20x speedup over naive encrypted implementations via MPC-friendly encoders, multi-step training, and O(1) communication Top-k search.

  15. IR3DE: A Linear Router for Large Language Models

    cs.CL 2026-06 unverdicted novelty 5.0

    IR3DE is a ridge regression router for domain-expert LLMs that matches or exceeds baselines in language modeling and reasoning tasks while allowing dynamic expert addition or removal without retraining.

  16. A Scalable Multi-LLM Collaboration System with Retrieval-based Selection and Exploration-Exploitation-Driven Enhancement

    cs.CL 2025-07 unverdicted novelty 5.0

    SMCS coordinates 15 open-source LLMs via retrieval-based prior selection and exploration-exploitation posterior enhancement, outperforming GPT-4.1 by 5.36% and GPT-o3-mini by 5.28% on eight benchmarks.