Identifying and Mitigating Systemic Measurement Bias in Production LLM Inference Benchmarks

• URL: http://arxiv.org/abs/2605.24217v2
• Date: Tue, 26 May 2026 05:47:59 GMT
• Title: Identifying and Mitigating Systemic Measurement Bias in Production LLM Inference Benchmarks
• Abstract summary: We show that benchmarking utilities rely on single-process, asyncio-driven architectures that introduce fundamental client-side queuing bottlenecks.<n>We propose an unbiased, multi-process evaluation framework that effectively distributes client-side load, ensuring negligible queuing overhead.
• Score: 5.044646432723069
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: As Large Language Models (LLMs) transition from research environments to production deployments, evaluating their performance against strict Service Level Objectives (SLOs) has become critical. However, current evaluation methodologies suffer from severe measurement bias at scale. We demonstrate that widely used benchmarking utilities rely on single-process, asyncio-driven architectures that introduce fundamental client-side queuing bottlenecks under high concurrency. By modeling the benchmarking client as an $M/G/1$ queue, we mathematically demonstrate how the Python Global Interpreter Lock (GIL) artificially inflates Time to First Token (TTFT) and Time Per Output Token (TPOT) metrics as request rates scale. To resolve this systematic inaccuracy, we propose an unbiased, multi-process evaluation framework that effectively distributes client-side load, ensuring negligible queuing overhead. Furthermore, we formalize a composite metric, Normalized Time Per Output Token (NTPOT), to robustly amortize end-to-end latency, including prefill and scheduling delays across sequence lengths. Our empirical evaluation demonstrates that this methodology successfully isolates pure serving engine performance, enabling accurate, reproducible profiling of LLMs at production scales exceeding thousands of queries per second.

Related papers

• LLM-as-a-Verifier: A General-Purpose Verification Framework [74.40111651545979]
  We introduce LLM-as-a-Verifier, a general-purpose verification framework.<n>It provides fine-grained feedback for agentic tasks without requiring additional training.<n>It achieves state-of-the-art performance on several benchmarks.
  arXiv  Detail & Related papers  (2026-07-06T17:59:35Z)
• The Silent Hyperparameter: Quantifying the Impact of Inference Backends on LLM Reproducibility [4.514361164656055]
  We show that the choice of backend alone can shift benchmark scores by up to 16.6 percentage points.<n>We find this is driven by system-level optimizations like caching and graphs, custom kernels, and engine-specific defaults in logit processing.
  arXiv  Detail & Related papers  (2026-05-19T08:37:27Z)
• LiveFMBench: Unveiling the Power and Limits of Agentic Workflows in Specification Generation [75.05397479715576]
  Large language models (LLMs) and agents have shown promising progress, but their true capabilities and failure modes remain unclear.<n>We present the first systematic and contamination-aware study of LLM- and agent-based formal specification generation for C programs.
  arXiv  Detail & Related papers  (2026-05-02T11:31:33Z)
• SWE-TRACE: Optimizing Long-Horizon SWE Agents Through Rubric Process Reward Models and Heuristic Test-Time Scaling [13.335296846555204]
  Current pipelines are bottlenecked by unoptimized demonstration data, sparse execution rewards, and computationally prohibitive inference scaling.<n>We present SWE-TRACE, a unified framework optimizing the SWE agent lifecycle across data curation, reinforcement learning (RL), and test-time inference.<n> experiments on standard SWE benchmarks demonstrate that SWE-TRACE significantly advances the state-of-the-art, maximizing resolution rates while drastically reducing both token consumption inference latency.
  arXiv  Detail & Related papers  (2026-04-16T09:41:47Z)
• Chimera: Latency- and Performance-Aware Multi-agent Serving for Heterogeneous LLMs [62.17306142810532]
  We present Chimera, a predictive scheduling system for multi-agent workflow serving on heterogeneous LLM clusters.<n>Chime traces the best latency, reducing end-to-end latency by 1.2--2.4$times$ and improving task performance by 8.0-9.5 percentage points on average over competitive baselines.
  arXiv  Detail & Related papers  (2026-03-23T17:01:42Z)
• MCP-Universe: Benchmarking Large Language Models with Real-World Model Context Protocol Servers [86.00932417210477]
  We introduce MCP-Universe, the first comprehensive benchmark specifically designed to evaluate LLMs in realistic and hard tasks through interaction with real-world MCP servers.<n>Our benchmark encompasses 6 core domains spanning 11 different MCP servers: Location Navigation, Repository Management, Financial Analysis, 3D Design, Browser Automation, and Web Searching.<n>We find that even SOTA models such as GPT-5 (43.72%), Grok-4 (33.33%) and Claude-4.0-Sonnet (29.44%) exhibit significant performance limitations.
  arXiv  Detail & Related papers  (2025-08-20T13:28:58Z)
• Semi-Clairvoyant Scheduling of Speculative Decoding Requests to Minimize LLM Inference Latency [4.372762934308627]
  We propose a semi-clairvoyant request scheduling algorithm called Least-Attained/Perceived-Service for Speculative Decoding (LAPS-SD)<n>LAPS-SD can effectively minimize average inference latency by adaptively scheduling requests according to their features during decoding.<n>Experiments show that LAPS-SD reduces inference latency by approximately 39% compared to state-of-the-art scheduling methods.
  arXiv  Detail & Related papers  (2025-05-20T04:12:37Z)
• ELIS: Efficient LLM Iterative Scheduling System with Response Length Predictor [5.097511974401423]
  ELIS is a serving system for Large Language Models (LLMs) featuring an Iterative Shortest Remaining Time First (ISRTF) scheduler.<n>ISRTF scheduler efficiently manages inference tasks with the shortest remaining time.
  arXiv  Detail & Related papers  (2025-05-14T04:50:00Z)
• PRISM: Self-Pruning Intrinsic Selection Method for Training-Free Multimodal Data Selection [68.8373788348678]
  Visual instruction tuning adapts pre-trained Multimodal Large Language Models to follow human instructions.<n>PRISM is the first training-free framework for efficient visual instruction selection.<n>It reduces the end-to-end time for data selection and model tuning to just 30% of conventional pipelines.
  arXiv  Detail & Related papers  (2025-02-17T18:43:41Z)
• COrAL: Order-Agnostic Language Modeling for Efficient Iterative Refinement [80.18490952057125]
  Iterative refinement has emerged as an effective paradigm for enhancing the capabilities of large language models (LLMs) on complex tasks. We propose Context-Wise Order-Agnostic Language Modeling (COrAL) to overcome these challenges. Our approach models multiple token dependencies within manageable context windows, enabling the model to perform iterative refinement internally.
  arXiv  Detail & Related papers  (2024-10-12T23:56:19Z)
• Etalon: Holistic Performance Evaluation Framework for LLM Inference Systems [11.712948114304925]
  Large language models (LLMs) in production can incur substantial costs. We propose Etalon, a comprehensive performance evaluation framework that includes fluidity-index. We also evaluate various open-source platforms and model-as-a-service offerings using Etalon.
  arXiv  Detail & Related papers  (2024-07-09T16:13:26Z)
• How to Prune Your Language Model: Recovering Accuracy on the "Sparsity May Cry'' Benchmark [60.72725673114168]
  We revisit the question of accurate BERT-pruning during fine-tuning on downstream datasets. We propose a set of general guidelines for successful pruning, even on the challenging SMC benchmark.
  arXiv  Detail & Related papers  (2023-12-21T03:11:30Z)

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.