{
  "id": 10590,
  "url": "https://arxiv.org/abs/2607.13801v1",
  "title": "Traffic-Aware Randomized Smoothing for LLM-Based Network Intrusion Detection",
  "summary": "Large language model (LLM)-based intrusion detection systems (IDS) are increasingly studied for security monitoring, yet their robustness against feasible traffic manipulation remains largely empirical. We present Traffic-Aware Randomized Smoothing (TA-RS), a classifier-agnostic certified defense that injects Gaussian noise exclusively into the directly controllable (DC) subspace -- features a remote attacker can modify -- during both fine-tuning and certification, aligning the smoothing distrib",
  "authors": "Zhenpeng Li",
  "category": "research",
  "topics": "military-security",
  "orgs": null,
  "regions": null,
  "published_at": "2026-07-15T13:08:19.000Z",
  "fetched_at": "2026-07-16T05:10:56.605Z",
  "source_slug": "arxiv-ethics",
  "source_name": "arXiv",
  "source_homepage": "https://arxiv.org",
  "ethics_ai_record_url": "https://ethics.ai/record/10590",
  "original_url": "https://arxiv.org/abs/2607.13801v1",
  "evidence_status": "source-only",
  "attribution": "via ethics.ai"
}