paper-with-me

홈 › Papers

Sentra-Guard: A Real-Time Multilingual Defense Against Adversarial LLM Prompts

2025-10-26 · Md. Mehedi Hasan, Sk Tanzir Mehedi, Ziaur Rahman, Rafid Mostafiz, Md. Abir Hossain arxiv

This paper presents a real-time modular defense system named Sentra-Guard. The system detects and mitigates jailbreak and prompt injection attacks targeting large language models (LLMs). The framework uses a hybrid architecture with FAISS-indexed SBERT embedding representations that capture the semantic meaning of prompts, combined with fine-tuned transformer classifiers, which are machine learning models specialized for distinguishing between benign and adversarial language inputs. It identifies adversarial prompts in both direct and obfuscated attack vectors. A core innovation is the classifier-retriever fusion module, which dynamically computes context-aware risk scores that estimate how likely a prompt is to be adversarial based on its content and context. The framework ensures multilingual resilience with a language-agnostic preprocessing layer. This component automatically translates non-English prompts into English for semantic evaluation, enabling consistent detection across over 100 languages. The system includes a HITL feedback loop, where decisions made by the automated system are reviewed by human experts for continual learning and rapid adaptation under adversarial pressure. Sentra-Guard maintains an evolving dual-labeled knowledge base of benign and malicious prompts, enhancing detection reliability and reducing false positives. Evaluation results show a 99.96% detection rate (AUC = 1.00, F1 = 1.00) and an attack success rate (ASR) of only 0.004%. This outperforms leading baselines such as LlamaGuard-2 (1.3%) and OpenAI Moderation (3.7%). Unlike black-box approaches, Sentra-Guard is transparent, fine-tunable, and compatible with diverse LLM backends. Its modular design supports scalable deployment in both commercial and open-source environments. The system establishes a new state-of-the-art in adversarial LLM defense.

📄 PDF Abstract BibTeX arXiv:2510.22628

Code (0)

등록된 구현이 없습니다.

Tasks

Continual Learning

Similar Papers 제목 키워드 기반

JAILJUDGE: A Comprehensive Jailbreak Judge Benchmark with Multi-Agent Enhanced Explanation Evaluation Framework

2024-10-11 · Fan Liu, Yue Feng, Zhao Xu, Lixin Su 외

Despite advancements in enhancing LLM safety against jailbreak attacks, evaluating LLM defenses remains a challenge, with current methods often lacking explainability and generalization to complex scenarios, leading to i…

Multilingual Collaborative Defense for Large Language Models

2025-05-17 · Hongliang Li, Jinan Xu, Gengping Cui, Changhao Guan 외

The robustness and security of large language models (LLMs) has become a prominent research area. One notable vulnerability is the ability to bypass LLM safeguards by translating harmful queries into rare or underreprese…

SENTRA: Selected-Next-Token Transformer for LLM Text Detection

2025-09-15 · Mitchell Plyler, Yilun Zhang, Alexander Tuzhilin, Saoud Khalifah 외 arxiv

LLMs are becoming increasingly capable and widespread. Consequently, the potential and reality of their misuse is also growing. In this work, we address the problem of detecting LLM-generated text that is not explicitly …

Text Detection

SEALGuard: Safeguarding the Multilingual Conversations in Southeast Asian Languages for LLM Software Systems

2025-07-11 · Wenliang Shan, Michael Fu, Rui Yang, Chakkrit Tantithamthavorn arxiv

Safety alignment is critical for LLM-powered systems. While recent LLM-powered guardrail approaches such as LlamaGuard achieve high detection accuracy of unsafe inputs written in English (e.g., ``How to create a bomb?'')…

Defenses & Enablers For Skill Injection Attacks on Terminal Based Agents

2026-06-01 · Yoshinari Fujinuma, Varun Gangal, Traian Rebedea, Makesh Narsimhan Sreedhar 외 arxiv

Large language model (LLM) agents increasingly rely on reusable skills i.e. documents describing task-specific procedures. However, this introduces a new attack surface for agents to manage. We study two complementary di…