惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

钛媒体:引领未来商业与生活新知
钛媒体:引领未来商业与生活新知
博客园_首页
Vercel News
Vercel News
Last Week in AI
Last Week in AI
罗磊的独立博客
Cyber Security Advisories - MS-ISAC
Cyber Security Advisories - MS-ISAC
IT之家
IT之家
美团技术团队
U
Unit 42
Google DeepMind News
Google DeepMind News
P
Proofpoint News Feed
J
Java Code Geeks
V
V2EX
量子位
腾讯CDC
S
SegmentFault 最新的问题
The GitHub Blog
The GitHub Blog
G
Google Developers Blog
D
DataBreaches.Net
雷峰网
雷峰网
让小产品的独立变现更简单 - ezindie.com
让小产品的独立变现更简单 - ezindie.com
博客园 - 聂微东
L
LangChain Blog
C
Check Point Blog

cs.LG updates on arXiv.org

Memory-Guided Trust-Region Bayesian Optimization (MG-TuRBO) for High Dimensions EngageTriBoost: Predictive Modeling of User Engagement in Digital Mental Health Intervention Using Explainable Machine Learning Reservoir observer enhanced with residual calibration and attention mechanism Efficient RL Training for LLMs with Experience Replay Wireless Communication Enhanced Value Decomposition for Multi-Agent Reinforcement Learning Adversarial Sensor Errors for Safe and Robust Wind Turbine Fleet Control IKKA: Inversion Classification via Critical Anomalies for Robust Visual Servoing Adaptive Simulation Experiment for LLM Policy Optimization EvoLen: Evolution-Guided Tokenization for DNA Language Model Smartwatch-Based Sitting Time Estimation in Real-World Office Settings Structural Evaluation Metrics for SVG Generation via Leave-One-Out Analysis Loom: A Scalable Analytical Neural Computer Architecture Spectral Geometry of LoRA Adapters Encodes Training Objective and Predicts Harmful Compliance Finite-Sample Analysis of Nonlinear Independent Component Analysis:Sample Complexity and Identifiability Bounds How does Chain of Thought decompose complex tasks? Uncertainty-Aware Transformers: Conformal Prediction for Language Models Adaptive Candidate Point Thompson Sampling for High-Dimensional Bayesian Optimization Using Synthetic Data for Machine Learning-based Childhood Vaccination Prediction in Narok, Kenya Delve into the Applicability of Advanced Optimizers for Multi-Task Learning Bridging SFT and RL: Dynamic Policy Optimization for Robust Reasoning Multi-Agent Decision-Focused Learning via Value-Aware Sequential Communication Predictive Entropy Links Calibration and Paraphrase Sensitivity in Medical Vision-Language Models Efficient Hierarchical Implicit Flow Q-learning for Offline Goal-conditioned Reinforcement Learning Modality-Aware Zero-Shot Pruning and Sparse Attention for Efficient Multimodal Edge Inference The nextAI Solution to the NeurIPS 2023 LLM Efficiency Challenge Feature-Label Modal Alignment for Robust Partial Multi-Label Learning Integrated electro-optic attention nonlinearities for transformers Toward World Models for Epidemiology Tracing the Chain: Deep Learning for Stepping-Stone Intrusion Detection Batch Distillation Data for Developing Machine Learning Anomaly Detection Methods
Confidently Wrong: Severity-Aware Calibration of Prompt-I...
[Submitted on 21 Jun 2026] · 2026-06-23 · via cs.LG updates on arXiv.org

Computer Science > Cryptography and Security

arXiv:2606.22659 (cs)

[Submitted on 21 Jun 2026]

View PDF

Abstract:Prompt-injection detectors are deployed as guards: a model scores an input and a downstream system trusts or blocks it on that score. I study the confidence of these scores, not only their accuracy, when the attack distribution shifts away from the clean benchmark on which the operating point was chosen. I evaluate three released detectors, ProtectAI-v2 and two Prompt-Guard-2 checkpoints, at a single source-calibrated threshold that I freeze and transport across five shifts. I report a severity metric S, how confident a detector is on the attacks it misses, alongside the false-negative rate and discrimination. Across every shift and every detector, severity on the missed attacks stays between 0.99 and 1.00 while the false-negative rate ranges from 0.01 to 0.97: when these detectors miss, they miss with near-certainty. All three confidently pass indirect behavior-hijack injection, a blind spot unanimous across two vendors and a fourfold size range. Standard pooled calibration error does not register this; one detector it rates well-calibrated, at 0.06, is miscalibrated at 0.91 on the attacks alone. Run against live models, the missed injections leak the majority of working exploits, passing them at the rate they catch others. A controlled experiment traces the cause to content-keying rather than injection structure, an instruction-tuned model used as a judge shows the same hijack blind spot, and a black-box rewriter exploits the content-keying to manufacture working confident misses, most effectively on the most dangerous attack category. Code and data are public.

Submission history

From: Md Anas Biswas [view email]
[v1] Sun, 21 Jun 2026 20:30:12 UTC (1,101 KB)

Current browse context:

cs.CR

Bookmark

BibSonomy Reddit

Bibliographic Tools

Bibliographic and Citation Tools

Bibliographic Explorer Toggle

Code, Data, Media

Code, Data and Media Associated with this Article

Demos

Demos

Related Papers

Recommenders and Search Tools

About arXivLabs

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.