CyberIntel ⬡ News
★ Saved ◆ Cyber Reads
← Back ◬ AI & Machine Learning Apr 14, 2026

Jailbreaking the Matrix: Nullspace Steering for Controlled Model Subversion

arXiv Security Archived Apr 14, 2026 ✓ Full text saved

arXiv:2604.10326v1 Announce Type: new Abstract: Large language models remain vulnerable to jailbreak attacks -- inputs designed to bypass safety mechanisms and elicit harmful responses -- despite advances in alignment and instruction tuning. We propose Head-Masked Nullspace Steering (HMNS), a circuit-level intervention that (i) identifies attention heads most causally responsible for a model's default behavior, (ii) suppresses their write paths via targeted column masking, and (iii) injects a pe

Full text archived locally
✦ AI Summary · Claude Sonnet


    Computer Science > Cryptography and Security [Submitted on 11 Apr 2026] Jailbreaking the Matrix: Nullspace Steering for Controlled Model Subversion Vishal Pramanik, Maisha Maliha, Susmit Jha, Sumit Kumar Jha Large language models remain vulnerable to jailbreak attacks -- inputs designed to bypass safety mechanisms and elicit harmful responses -- despite advances in alignment and instruction tuning. We propose Head-Masked Nullspace Steering (HMNS), a circuit-level intervention that (i) identifies attention heads most causally responsible for a model's default behavior, (ii) suppresses their write paths via targeted column masking, and (iii) injects a perturbation constrained to the orthogonal complement of the muted subspace. HMNS operates in a closed-loop detection-intervention cycle, re-identifying causal heads and reapplying interventions across multiple decoding attempts. Across multiple jailbreak benchmarks, strong safety defenses, and widely used language models, HMNS attains state-of-the-art attack success rates with fewer queries than prior methods. Ablations confirm that nullspace-constrained injection, residual norm scaling, and iterative re-identification are key to its effectiveness. To our knowledge, this is the first jailbreak method to leverage geometry-aware, interpretability-informed interventions, highlighting a new paradigm for controlled model steering and adversarial safety circumvention. Subjects: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI) Cite as: arXiv:2604.10326 [cs.CR]   (or arXiv:2604.10326v1 [cs.CR] for this version)   https://doi.org/10.48550/arXiv.2604.10326 Focus to learn more Submission history From: Maisha Maliha [view email] [v1] Sat, 11 Apr 2026 19:19:05 UTC (2,831 KB) Access Paper: HTML (experimental) view license Current browse context: cs.CR < prev   |   next > new | recent | 2026-04 Change to browse by: cs cs.AI References & Citations NASA ADS Google Scholar Semantic Scholar Export BibTeX Citation Bookmark Bibliographic Tools Bibliographic and Citation Tools Bibliographic Explorer Toggle Bibliographic Explorer (What is the Explorer?) Connected Papers Toggle Connected Papers (What is Connected Papers?) Litmaps Toggle Litmaps (What is Litmaps?) scite.ai Toggle scite Smart Citations (What are Smart Citations?) Code, Data, Media Demos Related Papers About arXivLabs Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)
    💬 Team Notes
    Article Info
    Source
    arXiv Security
    Category
    ◬ AI & Machine Learning
    Published
    Apr 14, 2026
    Archived
    Apr 14, 2026
    Full Text
    ✓ Saved locally
    Open Original ↗