CSGraph2Vec: Distributed Graph-Based Representation Learning for Assembly Functions
Bibliographic record
Abstract
Software reverse engineering is an essential but time-consuming undertaking in identifying malware, software vulnerabilities, and plagiarism, especially when access to the source code is limited. The extraction of abstract characteristics that represent malware and work as classifier inputs in traditional machine learning approaches requires feature engineering. The calibre of the features that are extracted has a major impact on how well these algorithms perform. In contrast, end-to-end learning solutions do not require hand-designed features and instead attempt to determine if an executable is harmful or not. However, a certain level of preprocessing remains essential in order to present malware content in a manner that the machine learning algorithm can comprehend. Due to the development of machine learning and deep learning, automating the construction of vector embeddings has become more feasible. This research introduces CSGraph2Vec, a distributed and automated deep learning approach that produces representations of assembly functions. Using the power of the Electra pre-trained language model, as well as message-passing neural networks, CSGraph2Vec efficiently incorporates control flow and semantic information from assembly code. Our model successfully learns significant features that distinguish benign from malicious functions. Through extensive experimentation and evaluation of the malware classification task, we show that our model performs better than several alternative approaches.
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 distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
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 teacher head, 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".