WireFall

Self-Hardening Transformer Web Application Firewall via Incremental Learning

• Built a firewall that catches zero-day attacks utilizing a DistilBERT encoder trained via Masked Language Modeling (MLM) for reconstruction loss, coupled with a frozen contrastive [CLS] head to detect zero-day anomalies with 93.78% recall.
• Red-Team self-hardening loop was designed where a fine-tuned Llama agent interprets high-loss anomalies to autonomously generate and inject regex signatures into the static engine, instantly patching new exploits.
• Acontinual learning framework was implemented to prevent catastrophic forgetting, leveraging Jensen-Shannon (JS) Divergence for stable knowledge distillation; derived a gradient-sensitivity metric to dynamically rank and stack LoRA adapters based on optimization relevance.