MalBERTv2: Code Aware BERT-Based Model for Malware Identification
Bibliographic record
Abstract
To proactively mitigate malware threats, cybersecurity tools, such as anti-virus and anti-malware software, as well as firewalls, require frequent updates and proactive implementation. However, processing the vast amounts of dataset examples can be overwhelming when relying solely on traditional methods. In cybersecurity workflows, recent advances in natural language processing (NLP) models can aid in proactively detecting various threats. In this paper, we present a novel approach for representing the relevance and significance of the Malware/Goodware (MG) datasets, through the use of a pre-trained language model called MalBERTv2. Our model is trained on publicly available datasets, with a focus on the source code of the apps by extracting the top-ranked files that present the most relevant information. These files are then passed through a pre-tokenization feature generator, and the resulting keywords are used to train the tokenizer from scratch. Finally, we apply a classifier using bidirectional encoder representations from transformers (BERT) as a layer within the model pipeline. The performance of our model is evaluated on different datasets, achieving a weighted f1 score ranging from 82% to 99%. Our results demonstrate the effectiveness of our approach for proactively detecting malware threats using NLP techniques.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.002 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.003 | 0.002 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".