סקור של השפעה ותפקוד של ראשי תשומת לב ב-Transformers: גילוי האפקטיביות של ראשי התשומת לב בזמן הפעלה

תקציר מקורי באנגליתarXiv:2609.05074v1 Announce Type: new Abstract: We propose an influence score to quantify the contribution of attention heads to classification decisions in Transformer-based models designed for prompt injection detection. The score combines directional influence on the logits with structural contribution within the residual stream, enabling a multi-scale analysis at the head, layer, and network levels. Applied to a DeBERTa model specialized for prompt injection detection, our framework reveals distinct decision behaviours between correct and erroneous predictions. Our method provides an effective compromise between fine-grained circuit analysis and global output-based methods, and offers a systematic way to study decision mechanisms in Transformer classifiers.