ReTabAD: A Benchmark for Restoring Semantic Context in Tabular Anomaly Detection

Dongmin Kim; Hankook Lee; Hye-Seung Cho; Sanghyu Yoon; Seungdong Yoa; Soonyoung Lee; Suhee Yoon; Woohyung Lim; Ye Seul Sim

REVIEW 3 major objections 6 minor 46 references

ReTabAD: A Benchmark for Restoring Semantic Context in Tabular Anomaly Detection

T0 review · 3 major / 6 minor · reviewed 2026-08-04 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read This paper claims that restoring textual metadata to tabular anomaly detection lets a zero-shot LLM reach accuracy comparable to trained state-of-the-art detectors, without any task-specific training.

desk verdict Real benchmark resource, but the headline gain conflates semantic context with telling the LLM which class is anomalous — the label-leakage control is missing. read the letter →

arxiv 2510.02060 v2 pith:3VXZ573E submitted 2025-10-02 cs.AI cs.LG

Sanghyu Yoon , Dongmin Kim , Suhee Yoon , Ye Seul Sim , Seungdong Yoa , Hye-Seung Cho , Soonyoung Lee , Hankook Lee

show 1 more author

Woohyung Lim

This is my paper

classification cs.AIcs.LG

keywords tabularanomalydetectionsemanticcontexttextualmetadatazero-shotLLMbenchmarkone-classclassificationinterpretabilityfeatureattribution

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Tabular anomaly detection benchmarks have traditionally stripped away the text that tells experts what a column means, what is normal, and what counts as an anomaly. This paper builds a 20-dataset benchmark that restores that textual metadata, pairing each dataset with structured JSON descriptions covering dataset background, per-column semantics, and label definitions. It then shows that a zero-shot LLM prompted with this metadata achieves an average AUROC of 0.847, matching or beating training-based detectors, while adding metadata raises AUROC by 7.6 points on average across five LLMs. The authors argue that semantic context improves both detection and interpretability, since the model can name and explain the features that drive a record's anomaly.

What carries the argument

The carrying object is the ReTabAD metadata file: a structured JSON per dataset containing a dataset-level description, per-column descriptions with logical types and units, and a label-level description stating which classes are normal and which are anomalous. This metadata is injected into a prompt template alongside per-column normal ranges (5th-95th percentiles) and analysis guidelines, and the LLM outputs anomaly scores, key features, and reasoning. The label-level description is the piece that lets the model map raw values directly onto anomaly semantics, and the paper also preserves raw numerical scales and restores categorical features to their original text values rather than arbitr

What would settle it

Run the zero-shot LLM in the full-description prompt with the label definition removed—keeping only dataset background, column semantics, and normal statistics—and compare AUROC. If removing the label definition on datasets like census or vertebral erases most of the reported gain, the result is driven by the told rule rather than by context-aware reasoning. A complementary test is to create a synthetic tabular dataset whose anomaly class is a rare pattern the metadata does not mention; the framework would predict only a small gain.

Watch

Extended reading notes

Core claim

The central discovery is that the definition of an anomaly is often carried by text, not just numbers, and a language model can use that text directly. The paper's zero-shot framework serializes each row with feature names and supplies three kinds of context—domain knowledge, feature descriptions, and normal statistics derived from the training data—then asks the LLM for an anomaly score, key features, and reasoning. With full metadata, the best-performing evaluator LLM reaches an average AUROC of 0.847 and an average rank of 4.08 among the 17 compared models, while the same model without descriptions scores 0.691. Ablations show that combining all three context types gives the highest win r

Load-bearing premise

The load-bearing premise is that the label-level descriptions in the metadata—which explicitly say which classes are normal and which are anomalous—are legitimate semantic context rather than leakage of the ground-truth anomaly rule; if that premise fails, the measured gains reflect rule-following, not context-aware detection.

Editorial extensions

If this is right

Semantic context can be treated as a first-class input in tabular anomaly detection, opening a benchmark standard for context-aware evaluation rather than purely numeric comparison.
A zero-shot LLM baseline becomes a strong reference point without task-specific training, lowering the cost of obtaining competitive detection baselines on new datasets.
Feature descriptions deliver large gains on categorical-heavy domains where raw integer codes are meaningless, so restoring textual categorical values can unlock performance that numeric-only pipelines miss.
Reasoning quality can be measured, and high-quality reasoning texts can be re-injected as few-shot examples to improve detection, giving an evaluation pathway for interpretability.
Anomaly definitions in benchmarks should be verified against original source documentation; several datasets in this benchmark differ from earlier processed versions after such verification.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

The label-level description in the prompt states the anomaly rule outright (e.g., high income is anomalous, or the original 'normal' class is anomalous). A fair reading is that the +7.6 AUROC gain measures how well LLMs apply a told rule; whether they can discover an anomaly rule from feature semantics alone is a testable open question.
If metadata becomes the bottleneck, automatic generation of column descriptions or retrieval from documentation could extend this style of context-aware detection to larger, uncurated datasets.
The framework points toward hybrid systems where statistical detectors flag candidates and LLMs re-rank or explain them using semantic context, rather than replacing trained detectors entirely.
The reasoning-alignment metric uses supervised attributions as ground truth; because those attributions come from a model trained on the same labels, high alignment could partly reflect agreement with the label rule rather than independent domain reasoning.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Two things to know up front. First, ReTabAD is a real resource: twenty tabular datasets with structured JSON metadata, source links, preprocessing scripts, and full appendix tables for sixteen baselines plus five LLMs under four prompt conditions. Anyone doing LLM-based tabular AD can now test whether textual metadata helps, and the documentation work is done properly. Second, the headline result — semantic context improves zero-shot detection by +7.6 AUROC and approaches trained SOTA — is overstated, because the 'domain knowledge' condition includes the Label-Level Description, which names the anomalous class outright.

The leakage is not subtle. In the paper's own words (§3.1), label descriptions specify 'which classes are considered normal and which are treated as anomalies.' The Type D prompt template (Figure 8) contains 'Target Label: {target_description}' and instructs the model to 'Leverage target label definition.' So the main ablation contrasts Type A (anonymized columns, statistics only) with Type D (statistics plus feature descriptions plus dataset background plus the explicit rule that this class is the anomaly). The trained baselines never get that rule. The +7.6 points, and the parity with MCM, therefore largely measure how well the LLM applies a told rule. To be fair: providing the label definition is defensible as a benchmark design choice, since practitioners usually know what they are hunting for. But then the claim needs reframing — this is not 'semantic context improves detection,' it is 'telling the model the target rule helps,' and the comparison should be against methods that also receive the target definition. The missing experiment is straightforward: full metadata with and without the label description. The paper does not run it, and it is the one control that would settle the question.

Secondary issues, in proportion. Some anomaly definitions are arbitrary rather than 'faithful': vertebral flips the original classes so the normal class is the anomaly, and census treats high income as anomalous. That weakens the curation claim, though the metadata and source links make the choices transparent and debatable. The F1@K reasoning-alignment metric uses SHAP from a label-trained XGBoost as ground truth; a reasonable proxy, and the paper says as much. The appendix is also honest about where alignment fails (quasar, covertype) and where sample sizes are too small (automobile). The Gemini Type C score (0.673, below Type A's 0.691) is a useful data point: domain text without statistics does not help, which the paper discusses.

Who this is for: anyone building or evaluating LLM-based tabular AD, and anyone who cares about benchmark hygiene in LLM evaluation. It deserves a serious referee — the flaw is fixable, the resource is valuable, the writing is transparent. I would send it to review with the label-leakage control required, and treat the headline number as unverified in the meantime.

Referee Report

3 major / 6 minor

Summary. ReTabAD introduces a benchmark of 20 tabular datasets enriched with structured textual metadata (dataset-level, column-level, and label-level descriptions), together with implementations of classical, deep, and LLM-based detectors and a zero-shot LLM baseline. The paper reports that adding semantic context to LLM prompts improves average AUROC by +7.6 points, that the zero-shot LLM achieves performance comparable to state-of-the-art training-based methods, and that metadata improves feature-attribution alignment and reasoning quality. The benchmark and code are publicly released.

Significance. If the central empirical claim were cleanly established, ReTabAD would be a useful community resource: it ships a metadata schema, baseline implementations, and a reproducible evaluation pipeline, and the idea of using feature/domain semantics in tabular AD is timely. However, the headline result is currently confounded because the 'Full Desc.' prompt contains the ground-truth anomaly rule through Label-Level Descriptions, so the zero-shot LLM is effectively told which class is anomalous. This moves the comparison from unsupervised context-aware AD toward rule application, and the interpretability analysis has an additional name-matching confound. With a proper control and appropriately narrowed claims, the benchmark could still be valuable; in its current form the main empirical claims are not yet supported.

major comments (3)

[§3.1, Figure 8, Tables 4–5] The central empirical claim is confounded by label leakage. §3.1 defines Label-Level Descriptions as specifying 'which classes are considered normal and which are treated as anomalies,' and the Type D prompt (Figure 8) includes 'Target Label: {target_description}' plus the instruction 'Leverage target label definition.' Thus the +7.6 AUROC gain (Table 4) and Type D win rates (Table 5) may measure the LLM's ability to apply a told rule, not the contribution of feature/domain semantics. Training-based unsupervised baselines in Table 3 never see this rule, making the comparison inequitable. Please add a control that compares Full Description with versus without the label description / target instruction, and report whether the gains persist without the label rule.
[§4.3, Appendix G, Table 6] The reasoning-alignment comparison is invalid as presented. Type A anonymizes column names (AA, AB, ...) while the SHAP reference set uses original feature names, and the paper does not describe any mapping between them. The near-zero F1@1/F1@3 values for Type A are therefore largely an artifact of name mismatch rather than missing semantics. Please either map predicted anonymized names back to original columns or evaluate Type A against anonymized references. Additionally, since Type D includes the target label definition, the reported F1 gains may again reflect rule following rather than improved semantic reasoning.
[§3.1, Table 8, Appendix A.1] The claim of 'faithful anomaly definitions' is undercut by several curation choices. Vertebral treats the original Normal class as anomalous and pathological classes as normal; census defines high income (>$50K) as anomalous; wine/glass select rare or less common classes. These are class-imbalance/rarity selections rather than domain-faithful anomalies. The criteria in §3.1 — discarding 'too easy' datasets and capping anomaly ratio at 1/3 — further distort representativeness. Please relabel these as 'class-defined evaluation labels' and provide robustness analyses, such as label-inversion sensitivity or anomaly-ratio sensitivity, to show that the benchmark's conclusions are not artifacts of these choices.

minor comments (6)

[Abstract, §3.2, Table 1] The manuscript inconsistently states 16 or 17 algorithms. Clarify whether the zero-shot LLM baseline is counted among the implemented algorithms or presented separately.
[Table 4] 'No Desc.' is a misnomer: Type A still provides statistical context (normal ranges/values). Rename it 'Statistical only' or 'No Metadata' to avoid implying that no textual information is used.
[Figure 2] The prompt structure explicitly shows 'Label 0 means Normal, 1 means Suspect/Pathologic.' If this label mapping is intentionally part of the semantic context, it should be disclosed and discussed as a source of supervision; if not, it should be removed from the baseline prompt.
[Tables 4–5, Appendix E] No statistical significance tests or confidence intervals are reported for average AUROC differences or win rates, despite multiple seeds and multiple LLMs. Add paired tests or bootstrap intervals for the headline +7.6 comparison.
[§4.2, Table 5] Type C includes the target label description, so the 'domain knowledge alone' condition is not clean. For the ablation to isolate domain knowledge, a variant without the label/target instruction is required; otherwise the Type A→D decomposition is uninterpretable.
[Title, Appendix D.3] Minor typos: 'Forrestoring' in the title, and '5th-9th percentile' in prompt templates appears to be a typo for '5th-95th percentile.' Also check consistency of hyphenation and spacing in 'Normal Ranges'.

Circularity Check

1 steps flagged · score 6.0 of 10

Zero-shot LLM 'semantic context' includes Label-Level Descriptions that encode the ground-truth anomaly class; the reported gains reduce, in part, to reading the told rule from the prompt.

self definitional [Section 3.1 (Label-Level Descriptions); Appendix D.3 Type D prompt (Figure 8); Table 8]
"Label-Level Descriptions: For each dataset, ReTabAD clearly specifies which classes are considered normal and which are treated as anomalies. These anomaly definitions are derived directly from the original dataset documentation... ## Target Label: {target_description} ... Leverage target label definition"

The 'semantic context' supplied to the zero-shot LLM includes Label-Level Descriptions, which are nothing other than the benchmark's ground-truth anomaly rule. The Type D prompt then presents 'Target Label: {target_description}' and explicitly instructs the model to 'Leverage target label definition.' Thus the LLM's anomaly score is, by construction, a function of the target label definition: the model is told which class is anomalous before it scores. The +7.6 AUROC improvement in Table 4 and the parity claim in Table 3 therefore measure rule-following from the prompt rather than context-aware anomaly detection. The comparison with training-based unsupervised methods is also inequitable, because those methods never see the target definition. A control with Full Desc. minus the label descr

full rationale

The paper's derivation chain for its central claim is: ReTabAD adds textual metadata -> zero-shot LLMs use that metadata -> semantic context improves detection performance and interpretability. The load-bearing step breaks because the metadata includes Label-Level Descriptions, defined as 'which classes are considered normal and which are treated as anomalies.' These same definitions are placed directly in the prompt as 'Target Label: {target_description}' with the instruction 'Leverage target label definition.' The LLM is therefore told the ground-truth anomaly class before producing scores. On datasets such as census, quasar, and vertebral, the anomaly rule is a simple class assertion (e.g., 'High-income individuals (>$50K) are considered anomalies,' 'Quasars (QSO) are considered anomalies,' 'Original Normal class samples are considered anomalies'), so the zero-shot LLM can solve the task by literal rule application. This makes the reported performance gains and comparisons partially circular: the input contains the answer by construction. The paper is not fully circular: the benchmark resource, feature-description-only ablations, and interpretability analyses have independent value, and there is no load-bearing self-citation chain. But the central empirical claim that 'semantic context improves detection performance' is confounded with label-rule leakage, warranting a score of 6 rather than 0 or 2.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

No free parameters are fitted in the zero-shot LLM; normal statistics are data-derived. The central claim rests on assumptions about what counts as semantic context, the validity of curated anomaly labels, and the reference model for reasoning quality.

assumptions (4)

domain assumption The definition of an anomaly is context-dependent and can be restored from textual documentation.
Section 1 and §3.1 motivate the entire benchmark; no independent evidence establishes that metadata captures the 'true' anomaly semantics.
ad hoc to paper Label descriptions are valid semantic context rather than leakage of the ground-truth target.
§3.1 explicitly includes Label-Level Descriptions in metadata and Appendix D.3 Type D prompt instructs 'Leverage target label definition'; the paper's zero-shot/unsupervised framing depends on this being legitimate.
domain assumption XGBoost + SHAP trained on ground-truth labels is a valid reference for LLM reasoning quality.
§4.3 and Appendix G use this proxy for interpretability; no independent validation that SHAP top features correspond to human/expert reasoning.
ad hoc to paper Curation choices (discard 'too easy' datasets, cap anomaly ratio at 1/3, select classes as anomalies) preserve representative AD difficulty.
§3.1 and A.2 state these criteria; they may inflate apparent gains and are not standard external constraints.

how reviews work

0 comments

Cite this review

Pith. "Pith review of ReTabAD: A Benchmark for Restoring Semantic Context in Tabular Anomaly Detection." pith.science (2026). https://pith.science/paper/3VXZ573E

@misc{pith2026251002060,
  author       = {Pith},
  title        = {Pith review of: ReTabAD: A Benchmark for Restoring Semantic Context in Tabular Anomaly Detection},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/3VXZ573E}},
  note         = {Machine review of arXiv:2510.02060}
}

read the original abstract

In tabular anomaly detection (AD), textual semantics often carry critical signals, as the definition of an anomaly is closely tied to domain-specific context. However, existing benchmarks provide only raw data points without semantic context, overlooking rich textual metadata such as feature descriptions and domain knowledge that experts rely on in practice. This limitation restricts research flexibility and prevents models from fully leveraging domain knowledge for detection. ReTabAD addresses this gap by restoring textual semantics to enable context-aware tabular AD research. We provide (1) 20 carefully curated tabular datasets enriched with structured textual metadata, together with implementations of state-of-the-art AD algorithms including classical, deep learning, and LLM-based approaches, and (2) a zero-shot LLM framework that leverages semantic context without task-specific training, establishing a strong baseline for future research. Furthermore, this work provides insights into the role and utility of textual metadata in AD through experiments and analysis. Results show that semantic context improves detection performance and enhances interpretability by supporting domain-aware reasoning. These findings establish ReTabAD as a benchmark for systematic exploration of context-aware AD.

Figures

Figures reproduced from arXiv: 2510.02060 by the authors.

**Figure 1.** Overall Data Collection and Annotation Process of ReTabAD. We preserve semantically rich tabular values and augment them with structured textual metadata, enabling rigorous evaluation of context-aware AD. Overall, our work emphasizes the importance of semantic information (e.g., textual metadata) in AD for tabular data. We hope that this perspective, together with our new benchmark, will guide future research toward… view at source ↗

**Figure 2.** Zero-shot LLM Baseline Overivew. We design prompts to evaluate the role of semantic metadata by incorporating domain knowledge (Cdomain), feature descriptions (Cf eature), and normal statistics (Cstatistic). The LLM generates outputs including anomaly scores, key features, and anomaly reasoning. 3.2 ALGORITHMS To provide a comprehensive comparison, we evaluate 16 representative unsupervised AD models, with full algo… view at source ↗

**Figure 3.** Quantitative Evaluation of LLM Reasoning via Performance Gain. Results are reported on the representative cirrhosis dataset. Qualitative Analysis of Reasoning Text. To quantitatively explore reasoning quality, we measure detection AUROC improvement when reasoning texts of detected anomalies are provided as contextual examples in the prompt, prioritizing samples with higher anomaly scores. As shown in [PITH_FULL_I… view at source ↗

Figures from the paper (5 more)

**Figure 4.** Figure 4: Reasoning Text Examples on the cirrhosis dataset. Left: JSON example. Right: Comparison of Type A (numeric deviation only) and Type D (domain grounded explanation). 5 CONCLUSION AND FUTURE WORK We present ReTabAD, a benchmark that restores textual semantics for context…

**Figure 5.** Figure 5: Type A Prompt Examples Analyze the data for anomaly detection. ## Features: ### Numerical features: {feature_1 name}: {feature_1 description} {feature_2 name}: {feature_1 description} … ### Categorical features: {feature_K name} {feature_K description} ## Statistical C…

**Figure 6.** Figure 6: Type B Prompt Examples 21 [PITH_FULL_IMAGE:figures/full_fig_p021_6.png]

**Figure 7.** Figure 7: Type C Prompt Examples You are a senior {dataset_name} expert analyzing data for anomaly detection. ## Dataset: {dataset_name} {dataset_description} ## Target Label: {target_description} ## Features: ### Numerical features: {feature_1 name}: {feature_1 description} {fe…

**Figure 8.** Figure 8: Type D Prompt Examples 22 [PITH_FULL_IMAGE:figures/full_fig_p022_8.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

46 extracted references · 6 linked inside Pith

[1]

write newline

" write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION format.date year duplicate empty "emp...
[2]

Financial fraud detection applying data mining techniques: A comprehensive review from 2009 to 2019

Khaled Gubran Al-Hashedi and Pritheega Magalingam. Financial fraud detection applying data mining techniques: A comprehensive review from 2009 to 2019. Computer Science Review, 40: 0 100402, 2021

2009
[3]

Claude opus 4.1

Anthropic. Claude opus 4.1. https://www.anthropic.com/news/claude-opus-4-1, August 2025. Accessed: 2025-08-20

2025
[4]

Classification-based anomaly detection for general data

Liron Bergman and Yedid Hoshen. Classification-based anomaly detection for general data. arXiv preprint arXiv:2005.02359, 2020

arXiv 2005
[5]

Lof: identifying density-based local outliers

Markus M Breunig, Hans-Peter Kriegel, Raymond T Ng, and J \"o rg Sander. Lof: identifying density-based local outliers. In Proceedings of the 2000 ACM SIGMOD international conference on Management of data, pp.\ 93--104, 2000

2000
[6]

Campos and J

D. Campos and J. Bernardes. Cardiotocography . UCI Machine Learning Repository, 2000. DOI : https://doi.org/10.24432/C51S4N

doi:10.24432/c51s4n 2000
[7]

On the evaluation of unsupervised outlier detection: measures, datasets, and an empirical study

Guilherme O Campos, Arthur Zimek, J \"o rg Sander, Ricardo JGB Campello, Barbora Micenkov \'a , Erich Schubert, Ira Assent, and Michael E Houle. On the evaluation of unsupervised outlier detection: measures, datasets, and an empirical study. Data mining and knowledge discovery, 30 0 (4): 0 891--927, 2016

2016
[8]

Anomaly detection: A survey

Varun Chandola, Arindam Banerjee, and Vipin Kumar. Anomaly detection: A survey. ACM computing surveys (CSUR), 41 0 (3): 0 1--58, 2009

2009

Show all 46 references

[9]

Xgboost: A scalable tree boosting system

Tianqi Chen and Carlos Guestrin. Xgboost: A scalable tree boosting system. In Proceedings of the 22nd acm sigkdd international conference on knowledge discovery and data mining, pp.\ 785--794, 2016

2016
[10]

Gemini 2.5: Pushing the frontier with advanced reasoning, multimodality, long context, and next generation agentic capabilities

Gheorghe Comanici, Eric Bieber, Mike Schaekermann, Ice Pasupat, Noveen Sachdeva, Inderjit Dhillon, Marcel Blistein, Ori Ram, Dan Zhang, Evan Rosen, et al. Gemini 2.5: Pushing the frontier with advanced reasoning, multimodality, long context, and next generation agentic capabil...

2025 arXiv
[11]

Gpt-4.1 sets the standard in automated experiment design using novel python libraries

Nuno Fachada, Daniel Fernandes, Carlos M Fernandes, Bruno D Ferreira-Saraiva, and Jo \ a o P Matos-Carvalho. Gpt-4.1 sets the standard in automated experiment design using novel python libraries. arXiv preprint arXiv:2508.00033, 2025

2025
[12]

Deep learning for medical anomaly detection--a survey

Tharindu Fernando, Harshala Gammulle, Simon Denman, Sridha Sridharan, and Clinton Fookes. Deep learning for medical anomaly detection--a survey. ACM Computing Surveys (CSUR), 54 0 (7): 0 1--37, 2021

2021
[13]

Evaluating large language models on time series feature understanding: A comprehensive taxonomy and benchmark

Elizabeth Fons, Rachneet Kaur, Soham Palande, Zhen Zeng, Tucker Balch, Manuela Veloso, and Svitlana Vyetrenko. Evaluating large language models on time series feature understanding: A comprehensive taxonomy and benchmark. arXiv preprint arXiv:2404.16563, 2024

2024 arXiv
[14]

A comparative evaluation of unsupervised anomaly detection algorithms for multivariate data

Markus Goldstein and Seiichi Uchida. A comparative evaluation of unsupervised anomaly detection algorithms for multivariate data. PloS one, 11 0 (4): 0 e0152173, 2016

2016
[15]

Adbench: Anomaly detection benchmark

Songqiao Han, Xiyang Hu, Hailiang Huang, Minqi Jiang, and Yue Zhao. Adbench: Anomaly detection benchmark. Advances in neural information processing systems, 35: 0 32142--32159, 2022

2022
[16]

Tabllm: Few-shot classification of tabular data with large language models

Stefan Hegselmann, Alejandro Buendia, Hunter Lang, Monica Agrawal, Xiaoyi Jiang, and David Sontag. Tabllm: Few-shot classification of tabular data with large language models. In International conference on artificial intelligence and statistics, pp.\ 5549--5581. PMLR, 2023

2023
[17]

Comparative analysis of machine learning models for anomaly detection in manufacturing

Andrey Kharitonov, Abdulrahman Nahhas, Matthias Pohl, and Klaus Turowski. Comparative analysis of machine learning models for anomaly detection in manufacturing. Procedia Computer Science, 200: 0 1288--1297, 2022

2022
[18]

Large language models are zero-shot reasoners

Takeshi Kojima, Shixiang Shane Gu, Machel Reid, Yutaka Matsuo, and Yusuke Iwasawa. Large language models are zero-shot reasoners. Advances in neural information processing systems, 35: 0 22199--22213, 2022

2022
[19]

Deep learning for anomaly detection in log data: A survey

Max Landauer, Sebastian Onder, Florian Skopik, and Markus Wurzenberger. Deep learning for anomaly detection in log data: A survey. Machine Learning with Applications, 12: 0 100470, 2023

2023
[20]

Rca: A deep collaborative autoencoder approach for anomaly detection

Boyang Liu, Ding Wang, Kaixiang Lin, Pang-Ning Tan, and Jiayu Zhou. Rca: A deep collaborative autoencoder approach for anomaly detection. In IJCAI: proceedings of the conference, volume 2021, pp.\ 1505, 2021

2021
[21]

Isolation forest

Fei Tony Liu, Kai Ming Ting, and Zhi-Hua Zhou. Isolation forest. In 2008 eighth ieee international conference on data mining, pp.\ 413--422. IEEE, 2008

2008
[22]

On diffusion modeling for anomaly detection

Victor Livernoche, Vineet Jain, Yashar Hezaveh, and Siamak Ravanbakhsh. On diffusion modeling for anomaly detection. arXiv preprint arXiv:2305.18593, 2023

2023 arXiv
[23]

Lundberg and Su - In Lee

Scott M. Lundberg and Su - In Lee. A unified approach to interpreting model predictions. In Proc. the Advances in Neural Information Processing Systems (NeurIPS), 2017

2017
[24]

Gpt-4o mini: Advancing cost-efficient intelligence

OpenAI . Gpt-4o mini: Advancing cost-efficient intelligence. https://openai.com/index/gpt-4o-mini-advancing-cost-efficient-intelligence/, 2024. Accessed: 2024-08-13

2024
[25]

Learning representations of ultrahigh-dimensional data for random distance-based outlier detection

Guansong Pang, Longbing Cao, Ling Chen, and Huan Liu. Learning representations of ultrahigh-dimensional data for random distance-based outlier detection. In Proceedings of the 24th ACM SIGKDD international conference on knowledge discovery & data mining, pp.\ 2041--2050, 2018
[26]

Deep learning for anomaly detection: A review

Guansong Pang, Chunhua Shen, Longbing Cao, and Anton Van Den Hengel. Deep learning for anomaly detection: A review. ACM computing surveys (CSUR), 54 0 (2): 0 1--38, 2021

2021
[27]

Neural transformation learning for deep anomaly detection beyond images

Chen Qiu, Timo Pfrommer, Marius Kloft, Stephan Mandt, and Maja Rudolph. Neural transformation learning for deep anomaly detection beyond images. In International conference on machine learning, pp.\ 8703--8714. PMLR, 2021

2021
[28]

Efficient algorithms for mining outliers from large data sets

Sridhar Ramaswamy, Rajeev Rastogi, and Kyuseok Shim. Efficient algorithms for mining outliers from large data sets. In Proceedings of the 2000 ACM SIGMOD international conference on Management of data, pp.\ 427--438, 2000

2000
[29]

Deep one-class classification

Lukas Ruff, Robert Vandermeulen, Nico Goernitz, Lucas Deecke, Shoaib Ahmed Siddiqui, Alexander Binder, Emmanuel M \"u ller, and Marius Kloft. Deep one-class classification. In International conference on machine learning, pp.\ 4393--4402. PMLR, 2018

2018
[30]

Kauffmann, Robert A

Lukas Ruff, Jacob R. Kauffmann, Robert A. Vandermeulen, Gr \' e goire Montavon, Wojciech Samek, Marius Kloft, Thomas G. Dietterich, and Klaus - Robert M \" u ller. A unifying review of deep and shallow anomaly detection. Proc. IEEE , 109 0 (5): 0 756--795, 2021 a

2021
[31]

A unifying review of deep and shallow anomaly detection

Lukas Ruff, Jacob R Kauffmann, Robert A Vandermeulen, Gr \'e goire Montavon, Wojciech Samek, Marius Kloft, Thomas G Dietterich, and Klaus-Robert M \"u ller. A unifying review of deep and shallow anomaly detection. Proceedings of the IEEE, 109 0 (5): 0 756--795, 2021 b

2021
[32]

The precision-recall plot is more informative than the roc plot when evaluating binary classifiers on imbalanced datasets

Takaya Saito and Marc Rehmsmeier. The precision-recall plot is more informative than the roc plot when evaluating binary classifiers on imbalanced datasets. PloS one, 10 0 (3): 0 e0118432, 2015

2015
[33]

Support vector method for novelty detection

Bernhard Sch \"o lkopf, Robert C Williamson, Alex Smola, John Shawe-Taylor, and John Platt. Support vector method for novelty detection. Advances in neural information processing systems, 12, 1999

1999
[34]

Estimating the support of a high-dimensional distribution

Bernhard Sch \"o lkopf, John C Platt, John Shawe-Taylor, Alex J Smola, and Robert C Williamson. Estimating the support of a high-dimensional distribution. Neural computation, 13 0 (7): 0 1443--1471, 2001

2001
[35]

A novel anomaly detection scheme based on principal component classifier

Mei-Ling Shyu, Shu-Ching Chen, Kanoksri Sarinnapakorn, and LiWu Chang. A novel anomaly detection scheme based on principal component classifier. Technical report, University of Miami, Department of Electrical and Computer Engineering, 2003

2003
[36]

Ano LLM : Large language models for tabular anomaly detection

Che-Ping Tsai, Ganyu Teng, Phillip Wallis, and Wei Ding. Ano LLM : Large language models for tabular anomaly detection. In The Thirteenth International Conference on Learning Representations, 2025. URL https://openreview.net/forum?id=7VkHffT5X2

2025
[37]

Unsupervised representation learning by predicting random distances

Hu Wang, Guansong Pang, Chunhua Shen, and Congbo Ma. Unsupervised representation learning by predicting random distances. arXiv preprint arXiv:1912.12186, 2019

1912 arXiv
[38]

Deep isolation forest for anomaly detection

Hongzuo Xu, Guansong Pang, Yijie Wang, and Yongjun Wang. Deep isolation forest for anomaly detection. IEEE Transactions on Knowledge and Data Engineering, 35 0 (12): 0 12591--12604, 2023 a . doi:10.1109/TKDE.2023.3270293

2023
[39]

Fascinating supervisory signals and where to find them: Deep anomaly detection with scale learning

Hongzuo Xu, Yijie Wang, Juhui Wei, Songlei Jian, Yizhou Li, and Ning Liu. Fascinating supervisory signals and where to find them: Deep anomaly detection with scale learning. In International Conference on Machine Learning, pp.\ 38655--38673. PMLR, 2023 b

2023
[40]

Qwen3 technical report

An Yang, Anfeng Li, Baosong Yang, Beichen Zhang, Binyuan Hui, Bo Zheng, Bowen Yu, Chang Gao, Chengen Huang, Chenxu Lv, et al. Qwen3 technical report. arXiv preprint arXiv:2505.09388, 2025

2025 arXiv
[41]

Drl: Decomposed representation learning for tabular anomaly detection

Hangting Ye, He Zhao, Wei Fan, Mingyuan Zhou, Dan dan Guo, and Yi Chang. Drl: Decomposed representation learning for tabular anomaly detection. In The Thirteenth International Conference on Learning Representations, 2025

2025
[42]

Mcm: Masked cell modeling for anomaly detection in tabular data

Jiaxin Yin, Yuanyuan Qiao, Zitang Zhou, Xiangchao Wang, and Jie Yang. Mcm: Masked cell modeling for anomaly detection in tabular data. In The Twelfth International Conference on Learning Representations, 2024

2024
[43]

Pyod: A python toolbox for scalable outlier detection

Yue Zhao, Zain Nasrullah, and Zheng Li. Pyod: A python toolbox for scalable outlier detection. Journal of machine learning research, 20 0 (96): 0 1--7, 2019

2019
[44]

@esa (Ref

\@ifxundefined[1] #1\@undefined \@firstoftwo \@secondoftwo \@ifnum[1] #1 \@firstoftwo \@secondoftwo \@ifx[1] #1 \@firstoftwo \@secondoftwo [2] @ #1 \@temptokena #2 #1 @ \@temptokena \@ifclassloaded agu2001 natbib The agu2001 class already includes natbib coding, so you should ...
[45]

\@lbibitem[] @bibitem@first@sw\@secondoftwo \@lbibitem[#1]#2 \@extra@b@citeb \@ifundefined br@#2\@extra@b@citeb \@namedef br@#2 \@nameuse br@#2\@extra@b@citeb \@ifundefined b@#2\@extra@b@citeb @num @parse #2 @tmp #1 NAT@b@open@#2 NAT@b@shut@#2 \@ifnum @merge>\@ne @bibitem@firs...
[46]

Record i : [feature\_name\_1=value\_1], [feature\_name\_2=value\_2],

@open @close @open @close and [1] URL: #1 \@ifundefined chapter * \@mkboth \@ifxundefined @sectionbib * \@mkboth * \@mkboth\@gobbletwo \@ifclassloaded amsart * \@ifclassloaded amsbook * \@ifxundefined @heading @heading NAT@ctr thebibliography [1] @ \@biblabel @NAT@ctr \@bibset...

1985

Pith tools

Reviewed August 4, 2026 · model on record in the stance chip above.

[1] [1]

write newline

" write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION format.date year duplicate empty "emp...

[2] [2]

Financial fraud detection applying data mining techniques: A comprehensive review from 2009 to 2019

Khaled Gubran Al-Hashedi and Pritheega Magalingam. Financial fraud detection applying data mining techniques: A comprehensive review from 2009 to 2019. Computer Science Review, 40: 0 100402, 2021

2009

[3] [3]

Claude opus 4.1

Anthropic. Claude opus 4.1. https://www.anthropic.com/news/claude-opus-4-1, August 2025. Accessed: 2025-08-20

2025

[4] [4]

Classification-based anomaly detection for general data

Liron Bergman and Yedid Hoshen. Classification-based anomaly detection for general data. arXiv preprint arXiv:2005.02359, 2020

arXiv 2005

[5] [5]

Lof: identifying density-based local outliers

Markus M Breunig, Hans-Peter Kriegel, Raymond T Ng, and J \"o rg Sander. Lof: identifying density-based local outliers. In Proceedings of the 2000 ACM SIGMOD international conference on Management of data, pp.\ 93--104, 2000

2000

[6] [6]

Campos and J

D. Campos and J. Bernardes. Cardiotocography . UCI Machine Learning Repository, 2000. DOI : https://doi.org/10.24432/C51S4N

doi:10.24432/c51s4n 2000

[7] [7]

On the evaluation of unsupervised outlier detection: measures, datasets, and an empirical study

Guilherme O Campos, Arthur Zimek, J \"o rg Sander, Ricardo JGB Campello, Barbora Micenkov \'a , Erich Schubert, Ira Assent, and Michael E Houle. On the evaluation of unsupervised outlier detection: measures, datasets, and an empirical study. Data mining and knowledge discovery, 30 0 (4): 0 891--927, 2016

2016

[8] [8]

Anomaly detection: A survey

Varun Chandola, Arindam Banerjee, and Vipin Kumar. Anomaly detection: A survey. ACM computing surveys (CSUR), 41 0 (3): 0 1--58, 2009

2009