Natural language information needs over symbolic music scores rarely reduce to a single-step lookup. Many queries require compositional Music Information Retrieval (MIR) that extracts multiple pieces of evidence from structured notation and aggregates them to answer the question. This setting remains challenging for Large Language Models due to the mismatch between natural language intents and symbolic representations, as well as the difficulty of reliably handling long structured contexts. Existing benchmarks only partially capture these retrieval demands, often emphasizing isolated theoretical knowledge or simplified settings. We introduce CSyMR-Bench, a benchmark for compositional MIR in symbolic music reasoning grounded in authentic user scenarios. It contains 126 multiple-choice questions curated from community discussions and professional examinations, where each item requires chaining multiple atomic analyses over a score to derive implicit musical evidence. To support diagnosis, we provide a taxonomy with six query intent categories and six analytical dimension tags. We further propose a tool-augmented retrieval and reasoning framework, CSyMR-Agent, that integrates a ReAct-style controller with deterministic symbolic analysis operators built with music21. Experiments across prompting baselines and agent variants show that tool-grounded compositional retrieval consistently outperforms Large Language Model-only approaches, yielding 5-7% absolute accuracy gains, with the largest improvements on analysis-heavy categories.

Summary Intrusion detection is essential to prevent damage to computer systems. However, in recent years, with the development of the network, many complex attack types have appeared, and it has become increasingly difficult to obtain high detection rates and low false alarm rates. In addition, traditional heavily hand‐crafted evaluation datasets for network intrusion detection have not been practical. This article proposes an intrusion detection method based on hierarchical feature learning, which can automatically learn traffic features. The method first learns the byte‐level features of network traffic through one‐dimensional convolutional neural networks and then learns session‐level features using stacked denoising autoencoder. The experiment analyzed the model structure and compared it with other methods. Experiments prove that the method in this article has high accuracy and low false alarm rate.