TimeSeriesExam - Anomaly Detection: leaderboard

Conceptual time-series understanding on TimeSeriesExam (2026), category anomaly detection (129 questions): multiple-choice questions on synthetic series built from expert-curated templates and refined with item response theory. Accuracy in percent; higher is better.

Metric: Accuracy (%, times 100) on the anomaly detection (129 questions) questions of TimeSeriesExam, 763 multiple-choice questions on synthetic series generated from over 100 expert-curated templates and refined over rounds with item response theory to separate candidate models; higher is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 4 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1Gemini 2.5 Pro59#276
2GPT-4o54#751
3Gemma 3 27B (IT)51#1324
4Qwen 2.5 VL 72B Instruct22#841

No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=timeseriesexam-anomaly-detection · How It Works · Data refreshed daily, snapshot 2026-11-21.