Benchmarking LLM Judges for Voice-Agent Evaluation: Reliability, Calibration, and Human Oversight

• URL: http://arxiv.org/abs/2608.24314v1
• Date: Tue, 25 Aug 2026 09:44:36 GMT
• Title: Benchmarking LLM Judges for Voice-Agent Evaluation: Reliability, Calibration, and Human Oversight
• Abstract summary: We compare human judgments with GPT-4.1 and GPT-5 on telecom and retail voice-agent conversations, across conversational quality and safety dimensions.<n>We examine metric-level correlations, evaluator consistency, and systematic human-LLM disagreement to identify which attributes can be judged reliably by automation.
• Score: 1.8147290789002162
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Evaluating conversational voice agents at scale re- quires reliable assessment methods that capture both observ- able interaction quality and the contextual judgment typically provided by human evaluators. We investigate LLM-as-a-Judge evaluation by comparing human judgments with GPT-4.1 and GPT-5 on telecom and retail voice-agent conversations, across conversational quality and safety dimensions. The same interac- tions are scored under three evaluation configurations, p0, p1, and p2, to test whether automated judgments are sensitive to the evaluation setup and whether observed patterns generalize across configurations and judge models. Beyond aggregate agreement, we examine metric-level correlations, evaluator consistency, and systematic human-LLM disagreement to identify which conver- sational attributes can be judged reliably by automation and which remain sensitive to interpretation and context. Effective voice-agent evaluation is also shaped by pipeline-level factors such as speech generation, streaming, and error propagation across ASR, reasoning, and tool-calling stages, motivating our focus on comparing how human and LLM judges score the same interactions end to end. Our results show that LLM- based evaluation can serve as an effective component of large- scale voice-agent assessment, but that its reliability is metric- and configuration-dependent rather than uniform. This pro- vides an empirical framework for identifying which metrics suit automated evaluation and supports hybrid pipelines in which LLM judges handle scalable assessment while human evaluators remain engaged for metrics that demand contextual interpretation and higher-confidence judgment.

Related papers

• Persona Conditioning as an Assessor-Sensitivity Probe for LLM-Based IR Evaluation [8.942079114981052]
  Large language models (LLMs) are increasingly used as relevance assessors in information retrieval (IR) evaluation.<n>We study persona conditioning as a diagnostic mechanism for exposing LLM assessor sensitivity.
  arXiv  Detail & Related papers  (2026-08-11T02:26:47Z)
• LLM-as-a-Judge for Evaluating System Responses in Conversational Music Recommendation [20.276417246845593]
  We present the first user study to empirically assess the reliability of LLM-as-a-judge for evaluating CRS responses.<n>We collect $n=400$ ratings from 20 domain-expert annotators, who evaluate each response across two dimensions: Personalization Quality and Explanation Quality.<n>We find that LLM-based judges exhibit moderate positive alignment with human assessments and outperform all reference-based baselines.
  arXiv  Detail & Related papers  (2026-07-28T12:26:55Z)
• Interactive ASR: Towards Human-Like Interaction and Semantic Coherence Evaluation for Agentic Speech Recognition [52.63676763985825]
  We propose a semantic-aware evaluation metric to assess recognition quality beyond token-level accuracy.<n>We design an agent framework to simulate human-like multi-turn interaction, enabling iterative refinement of recognition outputs.<n>We will release the code to facilitate future research in interactive and agentic ASR.
  arXiv  Detail & Related papers  (2026-04-10T09:02:42Z)
• How Trustworthy Are LLM-as-Judge Ratings for Interpretive Responses? Implications for Qualitative Research Workflows [0.6437935154416734]
  This study examines whether LLM-as-judge evaluations meaningfully align with human judgments of interpretive quality.<n>We generated one-sentence interpretive responses using five widely adopted inference models.<n>Results show that LLM-as-judge scores capture broad directional trends in human evaluations at the model level but diverge substantially in score magnitude.
  arXiv  Detail & Related papers  (2026-03-09T15:22:41Z)
• Pairwise or Pointwise? Evaluating Feedback Protocols for Bias in LLM-Based Evaluation [57.380464382910375]
  We show that the choice of feedback protocol for evaluation can significantly affect evaluation reliability and induce systematic biases.<n>We find that generator models can flip preferences by embedding distractor features.<n>We offer recommendations for choosing feedback protocols based on dataset characteristics and evaluation objectives.
  arXiv  Detail & Related papers  (2025-04-20T19:05:59Z)
• The simulation of judgment in LLMs [32.57692724251287]
  Large Language Models (LLMs) are increasingly embedded in evaluative processes, from information filtering to assessing and addressing knowledge gaps through explanation and credibility judgments.<n>This raises the need to examine how such evaluations are built, what assumptions they rely on, and how their strategies diverge from those of humans.<n>We benchmark six LLMs against expert ratings--NewsGuard and Media Bias/Fact Check--and against human judgments collected through a controlled experiment.
  arXiv  Detail & Related papers  (2025-02-06T18:52:10Z)
• HREF: Human Response-Guided Evaluation of Instruction Following in Language Models [61.273153125847166]
  We develop a new evaluation benchmark, Human Response-Guided Evaluation of Instruction Following (HREF)<n>In addition to providing reliable evaluation, HREF emphasizes individual task performance and is free from contamination.<n>We study the impact of key design choices in HREF, including the size of the evaluation set, the judge model, the baseline model, and the prompt template.
  arXiv  Detail & Related papers  (2024-12-20T03:26:47Z)
• Auto-PRE: An Automatic and Cost-Efficient Peer-Review Framework for Language Generation Evaluation [52.76508734756661]
  Auto-PRE is an automatic evaluation framework inspired by the peer review process.<n>Unlike previous approaches that rely on human annotations, Auto-PRE automatically selects evaluators based on three core traits.<n> Experiments on three representative tasks, including summarization, non-factoid QA, and dialogue generation, demonstrate that Auto-PRE achieves state-of-the-art performance.
  arXiv  Detail & Related papers  (2024-10-16T06:06:06Z)
• Aligning with Human Judgement: The Role of Pairwise Preference in Large Language Model Evaluators [48.54465599914978]
  Large Language Models (LLMs) have demonstrated promising capabilities as automatic evaluators in assessing the quality of generated natural language.<n>LLMs still exhibit biases in evaluation and often struggle to generate coherent evaluations that align with human assessments.<n>We introduce Pairwise-preference Search (PAIRS), an uncertainty-guided search-based rank aggregation method that employs LLMs to conduct pairwise comparisons locally and efficiently ranks candidate texts globally.
  arXiv  Detail & Related papers  (2024-03-25T17:11:28Z)
• Perspectives on Large Language Models for Relevance Judgment [56.935731584323996]
  Large language models (LLMs) claim that they can assist with relevance judgments. It is not clear whether automated judgments can reliably be used in evaluations of retrieval systems.
  arXiv  Detail & Related papers  (2023-04-13T13:08:38Z)

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.