SteerProbe: Learning to Bypass Safety Steering in Vision-Language Models

arXiv:2609.39117 · cs.CR · Submitted 2026-09-30 · Read on arXiv

cs.CR

Submitted: 2026-09-30

Updated: 2026-09-30

Terminology

• Steering — 76× in this paper · explained in 2 episodes
• Calibration — 61× in this paper · explained in 7 episodes
• Benchmark — 33× in this paper · explained in 2 episodes
• Robustness — 27× in this paper · explained in 5 episodes
• Grounding — 9× in this paper · explained in 3 episodes
• Vision-language models — 7× in this paper · explained in 2 episodes
• Harmful Rate — 5× in this paper
• Accuracy — 4× in this paper · explained in 1 episode
• Activation Steering — 4× in this paper · explained in 4 episodes
• Alignment — 3× in this paper · explained in 1 episode
• Feature Extraction — 3× in this paper · explained in 2 episodes
• Fine-Tuning — 3× in this paper · explained in 5 episodes
• Ablation — 2× in this paper · explained in 1 episode
• Post-Selection — 2× in this paper · explained in 2 episodes
• Regularization — 2× in this paper · explained in 3 episodes
• Surrogate Model — 2× in this paper · explained in 2 episodes
• Adversarial Examples — 1× in this paper · explained in 2 episodes
• Adversarial Robustness — 1× in this paper · explained in 3 episodes
• Adversarial Transferability — 1× in this paper · explained in 1 episode
• Commentary — 1× in this paper · explained in 2 episodes
• Embeddings — 1× in this paper · explained in 2 episodes
• Failure Modes — 1× in this paper · explained in 3 episodes
• Generative AI — 1× in this paper · explained in 2 episodes
• Hyperparameters — 1× in this paper · explained in 2 episodes
• Instruction Tuning — 1× in this paper · explained in 2 episodes
• Multimodal Inputs — 1× in this paper · explained in 2 episodes
• Transferability — 1× in this paper · explained in 4 episodes

Sources

• Qwen3-VL Technical Report
• Qwen2.5-VL Technical Report
• The Llama 3 Herd of Models
• Steering Language Models With Activation Engineering
• Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution
• Grounding-Driven Attack: Improving Encoder-based Adversarial Transferability against Large Vision-Language Models

Related papers

• SoK: AI-Augmented Binary Reversing
• Relaxed Sender Anonymity for CBDC Interbank Settlement: A Zero-Knowledge Approach on Permissioned EVM
• Calibration-Family Overfit: Why Trusted Sabotage Monitors Don't Transfer Across Lineages
• Efficient Fuzzy PSI under One-Sided Assumptions
• Sealing the Audit-Runtime Gap for LLM Skills
• Token Composition: A Graph Based on EVM Logs