Skip to content

Hermes Katana v3.1

Latest

Choose a tag to compare

@claudlos claudlos released this 03 Jun 14:29
a2ad59b

A research + models release: the empirical paper, two purpose-trained origin-aware classifiers, and the Proving Ground harness behind them.

📄 Paper

Cross-Platform Transferability of Prompt Injection Attacks: Universal Attack Surfaces and an Origin-Aware Defense — under paper/ (self-contained LaTeX; build with pdflatex/bibtex or tectonic).

  • 2,363 adversarial sessions across 16 model/harness combinations on 5 platforms (from a 17,643-attack pool).
  • Preregistered & rejected: a permission-gated CLI was more vulnerable than a flat agent harness (40.8% vs 10.2%, +30.65 pp, p < 1e-6) — harness design shapes outcomes but does not dominate model alignment.
  • Vulnerability falls with scale but never to zero (4B models 48–95%; frontier 8–10%).
  • A leakage-audited, origin-aware DeBERTa-v3-large reaches 9-class macro F1 0.938 at 0.48% FPR; deployed as middleware it drives effective attacks 12.40% → 0.00% (n=10,774).
  • A token ablation shows the defense is origin-aware yet robust: the deployed MiniLM scanner uses the origin tag while the large model is content-driven/invariant; neither performs origin routing.

🤖 Models (Hugging Face, MIT)

Download via the artifact CLI: katana artifacts download v17_large / v17_minilm (commit-pinned, integrity-verified).

🔧 Toolkit

  • Package version → 3.1.0.
  • Registered the v17 models in the artifact registry; v15 runtime scanners unchanged.
  • Proving Ground .fp.jsonl double-count fixed; verified compatible with the latest Hermes Agent.

See CHANGELOG.md for the full list.