Stochasticity in Agentic Evaluations: Quantifying Inconsistency with Intraclass Correlation

• URL: http://arxiv.org/abs/2512.06710v1
• Date: Sun, 07 Dec 2025 07:58:13 GMT
• Title: Stochasticity in Agentic Evaluations: Quantifying Inconsistency with Intraclass Correlation
• Abstract summary: Intraclass Correlation Coefficient (ICC) is a metric from measurement science to characterize variance.<n>ICC decomposes observed variance into between-Query variance (task difficulty) and within-Query variance (agent inconsistency)<n>We demonstrate that ICC converges by n=8-16 trials for structured tasks and n>=32 for complex reasoning, enabling practitioners to set evidence-based resampling budgets.
• Score: 0.7418138958218443
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: As large language models become components of larger agentic systems, evaluation reliability becomes critical: unreliable sub-agents introduce brittleness into downstream system behavior. Yet current evaluation practice, reporting a single accuracy number from a single run, obscures the variance underlying these results, making it impossible to distinguish genuine capability improvements from lucky sampling. We propose adopting Intraclass Correlation Coefficient (ICC), a metric from measurement science, to characterize this variance. ICC decomposes observed variance into between-query variance (task difficulty) and within-query variance (agent inconsistency), highlighting whether reported results reflect true capability or measurement noise. We evaluated on GAIA (Levels 1-3, measuring agentic capabilities across varying reasoning complexity) and FRAMES (measuring retrieval and factuality across multiple documents). We found that ICC varies dramatically with task structure, with reasoning and retrieval tasks (FRAMES) exhibit ICC=0.4955-0.7118 across models, and agentic tasks (GAIA) exhibiting ICC=0.304-0.774 across models. For sub-agent replacement decisions in agentic systems, accuracy improvements are only trustworthy if ICC also improves. We demonstrate that ICC converges by n=8-16 trials for structured tasks and n>=32 for complex reasoning, enabling practitioners to set evidence-based resampling budgets. We recommend reporting accuracy alongside ICC and within-query variance as standard practice, and propose updated Evaluation Cards capturing these metrics. By making evaluation stability visible, we aim to transform agentic benchmarking from opaque leaderboard competition to trustworthy experimental science. Our code is open-sourced at https://github.com/youdotcom-oss/stochastic-agent-evals.

Related papers

• Conformal Changepoint Localization and Root Cause Analysis with Corrupted Observations [55.76952683833966]
  This paper proposes weighted CONCH (W-CONCH) and weighted CROC (W-CROC) to reduce confidence set size when data may be corrupted.<n>Experiments on image-based and real-world changepoint and root-cause benchmarks show that uncertainty-based weighting substantially reduces confidence set size.
  arXiv  Detail & Related papers  (2026-07-29T05:16:59Z)
• Multimodal Learning on Low-Quality Data with Conformal Predictive Self-Calibration [72.0672328514289]
  Multimodal learning often grapples with the challenge of low-quality data.<n>We propose a unified framework, termed Conformal Predictive Self-Calibration.<n>Our framework consistently outperforms existing state-of-the-art methods.
  arXiv  Detail & Related papers  (2026-05-05T14:48:52Z)
• Taming Actor-Observer Asymmetry in Agents via Dialectical Alignment [59.536125286960186]
  Multi-agent frameworks assigning specialized roles are increasingly adopted to enable self-reflection and mutual auditing.<n>We find it simultaneously induces a human-like cognitive bias known as Actor-Observer Asymmetry.<n>ReTAS guides agents to synthesize conflicting viewpoints into an objective consensus.
  arXiv  Detail & Related papers  (2026-04-21T15:05:58Z)
• CADMAS-CTX: Contextual Capability Calibration for Multi-Agent Delegation [0.0]
  CADMAS-CTX is a framework for contextual capability calibration.<n>A hierarchical contextual capability profile replaces static skill-level confidence with context-conditioned posteriors.<n>We empirically validate our method on GAIA and SWE-bench benchmarks.
  arXiv  Detail & Related papers  (2026-04-20T08:30:28Z)
• K$α$LOS finds Consensus: A Meta-Algorithm for Evaluating Inter-Annotator Agreement in Complex Vision Tasks [4.297070083645049]
  We propose K$$LOS, a unified meta-algorithm that generalizes the "Localization First" principle to dataset quality evaluation.<n>By resolving spatial correspondence before assessing agreement, our framework transforms complex-categorical problems into nominal reliability.
  arXiv  Detail & Related papers  (2026-03-28T08:54:05Z)
• Quantifying Divergence in Inter-LLM Communication Through API Retrieval and Ranking [0.0]
  Large language models (LLMs) increasingly operate as autonomous agents that reason over external APIs to perform complex tasks.<n>We present a unified benchmarking framework to quantify inter-LLM divergence, defined as the extent to which models differ in API discovery and ranking under identical tasks.
  arXiv  Detail & Related papers  (2026-03-09T20:28:24Z)
• ProtoDCS: Towards Robust and Efficient Open-Set Test-Time Adaptation for Vision-Language Models [32.840734752367275]
  Prototype-based Double-Check Separation (ProtoDCS) is a robust framework for OSTTA.<n>It separates csID and csOOD samples, enabling safe and efficient adaptation of Vision-Language Models to csID data.<n>ProtoDCS significantly boosts both known-class accuracy and OOD detection metrics.
  arXiv  Detail & Related papers  (2026-02-27T03:39:02Z)
• From Global to Granular: Revealing IQA Model Performance via Correlation Surface [83.65597122328133]
  We present textbfGranularity-Modulated Correlation (GMC), which provides a structured, fine-grained analysis of IQA performance.<n>GMC includes a textbfDistribution Regulator that regularizes correlations to mitigate biases from non-uniform quality distributions.<n>Experiments on standard benchmarks show that GMC reveals performance characteristics invisible to scalar metrics, offering a more informative and reliable paradigm for analyzing, comparing, and deploying IQA models.
  arXiv  Detail & Related papers  (2026-01-29T13:55:26Z)
• Gaming the Judge: Unfaithful Chain-of-Thought Can Undermine Agent Evaluation [76.5533899503582]
  Large language models (LLMs) are increasingly used as judges to evaluate agent performance.<n>We show this paradigm implicitly assumes that the agent's chain-of-thought (CoT) reasoning faithfully reflects both its internal reasoning and the underlying environment state.<n>We demonstrate that manipulated reasoning alone can inflate false positive rates of state-of-the-art VLM judges by up to 90% across 800 trajectories spanning diverse web tasks.
  arXiv  Detail & Related papers  (2026-01-21T06:07:43Z)
• Improving Consistency in Retrieval-Augmented Systems with Group Similarity Rewards [14.535325886547112]
  RAG systems are increasingly deployed in high-stakes domains where users expect outputs to be consistent across semantically equivalent queries.<n>Existing systems often exhibit significant inconsistencies due to variability in both the retriever and generator.<n>We introduce a principled evaluation framework that decomposes RAG consistency into retriever-level, generator-level, and end-to-end components.
  arXiv  Detail & Related papers  (2025-10-05T23:14:13Z)
• The Lie of the Average: How Class Incremental Learning Evaluation Deceives You? [48.83567710215299]
  Class Incremental Learning (CIL) requires models to continuously learn new classes without forgetting previously learned ones.<n>We argue that a robust CIL evaluation protocol should accurately characterize and estimate the entire performance distribution.<n>We propose EDGE, an evaluation protocol that adaptively identifies and samples extreme class sequences using inter-task similarity.
  arXiv  Detail & Related papers  (2025-09-26T17:00:15Z)
• Cumulative Consensus Score: Label-Free and Model-Agnostic Evaluation of Object Detectors in Deployment [3.6178660238507843]
  evaluating object detection models in deployment is challenging because ground-truth annotations are rarely available.<n>We introduce the Cumulative Consensus Score (CCS), a label-free metric that enables continuous monitoring and comparison of detectors in real-world settings.
  arXiv  Detail & Related papers  (2025-09-16T09:24:37Z)
• Collaborative Value Function Estimation Under Model Mismatch: A Federated Temporal Difference Analysis [55.13545823385091]
  Federated reinforcement learning (FedRL) enables collaborative learning while preserving data privacy by preventing direct data exchange between agents.<n>In real-world applications, each agent may experience slightly different transition dynamics, leading to inherent model mismatches.<n>We show that even moderate levels of information sharing significantly mitigate environment-specific errors.
  arXiv  Detail & Related papers  (2025-03-21T18:06:28Z)
• GMC-IQA: Exploiting Global-correlation and Mean-opinion Consistency for No-reference Image Quality Assessment [40.33163764161929]
  We construct a novel loss function and network to exploit Global-correlation and Mean-opinion Consistency. We propose a novel GCC loss by defining a pairwise preference-based rank estimation to solve the non-differentiable problem of SROCC. We also propose a mean-opinion network, which integrates diverse opinion features to alleviate the randomness of weight learning.
  arXiv  Detail & Related papers  (2024-01-19T06:03:01Z)
• Diagnosing and Rectifying Fake OOD Invariance: A Restructured Causal Approach [51.012396632595554]
  Invariant representation learning (IRL) encourages the prediction from invariant causal features to labels de-confounded from the environments. Recent theoretical results verified that some causal features recovered by IRLs merely pretend domain-invariantly in the training environments but fail in unseen domains. We develop an approach based on conditional mutual information with respect to RS-SCM, then rigorously rectify the spurious and fake invariant effects.
  arXiv  Detail & Related papers  (2023-12-15T12:58:05Z)
• Newer is not always better: Rethinking transferability metrics, their peculiarities, stability and performance [5.650647159993238]
  Fine-tuning of large pre-trained image and language models on small customized datasets has become increasingly popular. We show that the statistical problems with covariance estimation drive the poor performance of H-score. We propose a correction and recommend measuring correlation performance against relative accuracy in such settings.
  arXiv  Detail & Related papers  (2021-10-13T17:24:12Z)
• Learning Calibrated Uncertainties for Domain Shift: A Distributionally Robust Learning Approach [150.8920602230832]
  We propose a framework for learning calibrated uncertainties under domain shifts. In particular, the density ratio estimation reflects the closeness of a target (test) sample to the source (training) distribution. We show that our proposed method generates calibrated uncertainties that benefit downstream tasks.
  arXiv  Detail & Related papers  (2020-10-08T02:10:54Z)
• Meta-Learned Confidence for Few-shot Learning [60.6086305523402]
  A popular transductive inference technique for few-shot metric-based approaches, is to update the prototype of each class with the mean of the most confident query examples. We propose to meta-learn the confidence for each query sample, to assign optimal weights to unlabeled queries. We validate our few-shot learning model with meta-learned confidence on four benchmark datasets.
  arXiv  Detail & Related papers  (2020-02-27T10:22:17Z)

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.