Text to be classified
BERT + Statistical Features
6 Specialized Models
Majority Decision
Human vs AI-generated
Leverages pre-trained models' semantic understanding capabilities:
Integrates multiple linguistic features:
Probabilistic analysis methods:
Figure 1: Detection accuracy comparison showing EBF Detection outperforming other methods with 98.55% accuracy
Figure 2: Multi-Feature Detection Components showing the integration of token probability, syntactic structure, and lexical diversity features
Figure 3: Radar chart comparing in-domain and out-of-domain performance across different detection methods
Figure 4: Overall architecture of EBF Detection showing the voting mechanism combining six detectors
Method | In-Domain Accuracy (%) | Out-of-Domain Accuracy (%) | F1 Score |
---|---|---|---|
Log-Probability | 90.19 | 89.77 | 90.42 |
Log-Rank | 89.61 | 90.40 | 90.32 |
GLTR | 90.17 | 88.10 | 90.14 |
DetectGPT | 79.26 | 87.52 | 79.69 |
BERT | 90.58 | 85.66 | 90.60 |
EBF Detection | 98.55 | 96.06 | 98.47 |
The BERT Detector and BERTTextCNN leverage pre-trained models' semantic understanding capabilities, which are effective in capturing deep contextual differences between human and AI-generated text. These models excel at understanding semantic coherence and contextual relationships.
Multi-Feature Detection focuses on shallow linguistic statistics including:
The integration process combines outputs using majority voting, leveraging the complementary strengths of deep semantic features (from BERT-based models) and shallow statistical features (from multi-feature and log-based detectors). This ensures comprehensive coverage of both semantic and stylistic differences.
This paper introduces EBF Detection, a robust method for detecting text generated by large language models. By integrating multiple weak detectors through ensemble learning, the approach achieves:
Industry-leading accuracy of 98.72% on in-domain data, significantly outperforming existing methods
Maintains 96.79% accuracy on out-of-domain data, demonstrating robustness to domain shifts
Does not require watermarking or modification of generated text, preserving naturalness
Future research will focus on: