SMU_MalwareDetection: Assembly Code Fragments from Malicious and Benign PE Files

Assembly code fragments extracted from malicious and benign Portable Executable (PE) files. The dataset was created to support deep learning-based malware classification and is hosted on Hugging Face by author qwewerewq. It was last updated on June 13, 2026.

Use Cases

• Train malware classifiers based on assembly code fragments.
• Benchmark deep learning models for security applications using segmented code.
• Study contextual patterns in malicious versus benign executable code.
• Investigate class imbalance challenges in security datasets.

Strengths

• Assembly code was segmented using a sliding window technique to preserve contextual information.
• Explicitly notes a class imbalance between benign (label 0) and malware (label 1) samples.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

qwewerewq on Hugging Face

Collection Method

Extracted from Portable Executable (PE) files and segmented with a sliding window.

Freshness

Last updated 2026-06-13 22:50:50; freshness should be verified.