Payload-Aware Intrusion Detection with CMAE and Large Language Models
Bibliographic record
Abstract
Intrusion Detection Systems (IDS) play a vital role in network security, yet signature-based methods are limited by high false positive rates (FPR) and inability to detect novel threats. Recent AI-based approaches offer improved adaptability, but most rely on flow-level or statistical features, constraining their ability to analyze sophisticated payload-based attacks. To address these challenges, we present a dual-path IDS framework: Xavier-CMAE, a lightweight model using Hex2Int tokenization and Xavier initialization, achieves 99.9718% accuracy and a 0.0182% FPR without pre-training; and LLM-CMAE, which leverages pre-trained LLM tokenizers for enhanced detection, achieves 99.9696% accuracy and a 0.0194% FPR at higher computational cost. Experimental results on the CIC-IDS2017 dataset reveal a distinct trade-off between efficiency and Contextually Adept and Scalable (CAS) power, indicating that a modular approach may enable both real-time scalability and in-depth threat analysis. This work advances AI-powered intrusion detection by (1) introducing a modular, payload-centric dual-path architecture that combines lightweight and CAS detection for adaptive, layered security; (2) demonstrating that Xavier-CMAE achieves real-time scalability and state-of-the-art accuracy without embedding pre-training; and (3) exploring the effectiveness and future potential of integrating pre-trained LLM tokenizers for nuanced, selective threat analysis and robust IDS design.
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 | 0.002 | 0.006 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.002 | 0.004 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.001 | 0.003 |
| Insufficient payload (model declined to judge) | 0.003 | 0.003 |
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".