An Empirical Study of SOTA RCA Models: From Oversimplified Benchmarks to Realistic Failures

• URL: http://arxiv.org/abs/2510.04711v1
• Date: Mon, 06 Oct 2025 11:30:03 GMT
• Title: An Empirical Study of SOTA RCA Models: From Oversimplified Benchmarks to Realistic Failures
• Authors: Aoyang Fang, Songhan Zhang, Yifan Yang, Haotong Wu, Junjielong Xu, Xuyang Wang, Rui Wang, Manyi Wang, Qisheng Lu, Pinjia He,
• Abstract summary: We show that simple rule-based methods can match or even outperform state-of-the-art (SOTA) models on four widely used benchmarks.<n>Our analysis highlights three common failure patterns: scalability issues, observability blind spots, and modeling bottlenecks.
• Score: 16.06503310632004
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/
• Abstract: While cloud-native microservice architectures have transformed software development, their complexity makes Root Cause Analysis (RCA) both crucial and challenging. Although many data-driven RCA models have been proposed, we find that existing benchmarks are often oversimplified and fail to capture real-world conditions. Our preliminary study shows that simple rule-based methods can match or even outperform state-of-the-art (SOTA) models on four widely used benchmarks, suggesting performance overestimation due to benchmark simplicity. To address this, we systematically analyze popular RCA benchmarks and identify key limitations in fault injection, call graph design, and telemetry patterns. Based on these insights, we develop an automated framework to generate more realistic benchmarks, yielding a dataset of 1,430 validated failure cases from 9,152 injections, covering 25 fault types under dynamic workloads with hierarchical ground-truth labels and verified SLI impact. Re-evaluation of 11 SOTA models on this dataset shows low Top@1 accuracy (average 0.21, best 0.37) and significantly longer execution times. Our analysis highlights three common failure patterns: scalability issues, observability blind spots, and modeling bottlenecks.

Related papers

• STAR : Bridging Statistical and Agentic Reasoning for Large Model Performance Prediction [78.0692157478247]
  We propose STAR, a framework that bridges data-driven STatistical expectations with knowledge-driven Agentic Reasoning.<n>We show that STAR consistently outperforms all baselines on both score-based and rank-based metrics.
  arXiv  Detail & Related papers  (2026-02-12T16:30:07Z)
• Evaluating and Enhancing the Vulnerability Reasoning Capabilities of Large Language Models [15.849480549367684]
  We propose DAGVul, a novel framework that models vulnerability reasoning as a Directed Acyclic Graph (DAG) generation task.<n>By further introducing Reinforcement Learning with Verifiable Rewards (RLVR), we align model reasoning trace with program-intrinsic logic.<n>Our framework improves the reasoning F1-score by an average of 18.9% over all the baselines.
  arXiv  Detail & Related papers  (2026-02-06T13:19:45Z)
• Simplicity Prevails: The Emergence of Generalizable AIGI Detection in Visual Foundation Models [15.709482146201283]
  A simple linear classifier trained on the frozen features of modern Vision Foundation Models establishes a new state-of-the-art.<n>We show that this baseline matches specialized detectors on standard benchmarks but also decisively outperforms them on in-the-wild datasets.<n>We conclude by advocating for a paradigm shift in AI forensics, moving from overfitting on static benchmarks to harnessing the evolving world knowledge of foundation models for real-world reliability.
  arXiv  Detail & Related papers  (2026-02-02T07:20:02Z)
• How Reliable are Confidence Estimators for Large Reasoning Models? A Systematic Benchmark on High-Stakes Domains [7.845652284569666]
  miscalibration of Large Reasoning Models undermines their reliability in high-stakes domains.<n>We introduce the Reasoning Model Confidence estimation Benchmark (RMCB), a public resource of 347,496 reasoning traces from six popular LRMs.
  arXiv  Detail & Related papers  (2026-01-13T01:55:48Z)
• Labels Matter More Than Models: Quantifying the Benefit of Supervised Time Series Anomaly Detection [56.302586730134806]
  Time series anomaly detection (TSAD) is a critical data mining task often constrained by label scarcity.<n>Current research predominantly focuses on Unsupervised Time-series Anomaly Detection.<n>This paper challenges the premise that architectural complexity is the optimal path for TSAD.
  arXiv  Detail & Related papers  (2025-11-20T08:32:49Z)
• Beyond Model Ranking: Predictability-Aligned Evaluation for Time Series Forecasting [18.018179328110048]
  We introduce a predictability-aligned diagnostic framework grounded in spectral coherence.<n>We provide the first systematic evidence of "predictability drift", demonstrating that a task's forecasting difficulty varies sharply over time.<n>Our evaluation reveals a key architectural trade-off: complex models are superior for low-predictability data, whereas linear models are highly effective on more predictable tasks.
  arXiv  Detail & Related papers  (2025-09-27T02:56:06Z)
• Rethinking Evaluation of Infrared Small Target Detection [105.59753496831739]
  This paper introduces a hybrid-level metric incorporating pixel- and target-level performance, proposing a systematic error analysis method, and emphasizing the importance of cross-dataset evaluation.<n>An open-source toolkit has be released to facilitate standardized benchmarking.
  arXiv  Detail & Related papers  (2025-09-21T02:45:07Z)
• RoHOI: Robustness Benchmark for Human-Object Interaction Detection [78.18946529195254]
  Human-Object Interaction (HOI) detection is crucial for robot-human assistance, enabling context-aware support.<n>We introduce the first benchmark for HOI detection, evaluating model resilience under diverse challenges.<n>Our benchmark, RoHOI, includes 20 corruption types based on the HICO-DET and V-COCO datasets and a new robustness-focused metric.
  arXiv  Detail & Related papers  (2025-07-12T01:58:04Z)
• SHARDeg: A Benchmark for Skeletal Human Action Recognition in Degraded Scenarios [3.0519884745675485]
  Skeletal Human Action Recognition (SHAR) is critical in many CV pipelines operating in real-time and at the edge.<n>This is exemplified by Skeletal Human Action Recognition (SHAR), which is critical in many CV pipelines operating in real-time and at the edge.<n>We demonstrate the need for this benchmark by showing that the form of degradation, which has not previously been considered, has a large impact on model accuracy.<n>We identify that temporal regularity of frames in degraded SHAR data is likely a major driver of differences in model performance.
  arXiv  Detail & Related papers  (2025-05-23T15:52:31Z)
• Towards Robust Universal Information Extraction: Benchmark, Evaluation, and Solution [66.11004226578771]
  Existing robust benchmark datasets have two key limitations.<n>They generate only a limited range of perturbations for a single Information Extraction (IE) task.<n>Considering the powerful generation capabilities of Large Language Models (LLMs), we introduce a new benchmark dataset for Robust UIE, called RUIE-Bench.<n>We show that training with only textbf15% of the data leads to an average textbf7.5% relative performance improvement across three IE tasks.
  arXiv  Detail & Related papers  (2025-03-05T05:39:29Z)
• Optimizing Sequential Recommendation Models with Scaling Laws and Approximate Entropy [104.48511402784763]
  Performance Law for SR models aims to theoretically investigate and model the relationship between model performance and data quality.<n>We propose Approximate Entropy (ApEn) to assess data quality, presenting a more nuanced approach compared to traditional data quantity metrics.
  arXiv  Detail & Related papers  (2024-11-30T10:56:30Z)
• PerturBench: Benchmarking Machine Learning Models for Cellular Perturbation Analysis [14.298235969992877]
  We introduce a comprehensive framework for perturbation response modeling in single cells.<n>Our approach includes a modular and user-friendly model development and evaluation platform.<n>We highlight the limitations of widely used models, such as mode collapse.
  arXiv  Detail & Related papers  (2024-08-20T07:40:20Z)
• QualEval: Qualitative Evaluation for Model Improvement [82.73561470966658]
  We propose QualEval, which augments quantitative scalar metrics with automated qualitative evaluation as a vehicle for model improvement. QualEval uses a powerful LLM reasoner and our novel flexible linear programming solver to generate human-readable insights. We demonstrate that leveraging its insights, for example, improves the absolute performance of the Llama 2 model by up to 15% points relative.
  arXiv  Detail & Related papers  (2023-11-06T00:21:44Z)

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.