Comparison of Meta-Learning Convergence and F1-Score in Small and Large Language Models for Few-Shot Anomaly Detection

Anomaly detection is a widely explored domain in machine learning. Many models are proposed in the literature, and compared through different metrics measured on various datasets. The most popular metrics used to compare performances are F1-score, AUC and AVPR. In this paper, we show that F1-score and AVPR are highly sensitive to the contamination rate. One consequence is that it is possible to artificially increase their values by modifying the train-test split procedure. This leads to misleading comparisons between algorithms in the literature, especially when the evaluation protocol is not

Research goal: How does the convergence speed and final F1-score of meta-learning frameworks compare between small (8B) and large (70B) language models in cross-domain few-shot anomaly detection?

Autonomous synthesis report generated by SOVEREIGN Research Kernel. Tribunal consensus score: 7.5/10.