Copertina

When Good Sounds Go Adversarial: Jailbreaking Audio-Language Models with Benign Inputs

Fuente: arXiv

Salvato in:

Dettagli Bibliografici
Autori principali:	Dingeto, Hiskias, Kwon, Taeyoun, Choi, Dasol, Kim, Bodam, Lee, DongGeon, Park, Haon, Lee, JaeHoon, Shin, Jongho
Natura:	Preprint
Pubblicazione:	2025
Soggetti:	Sound Artificial Intelligence Cryptography and Security Audio and Speech Processing
Accesso online:	Acceder al recurso
Tags:	Aggiungi Tag Nessun Tag, puoi essere il primo ad aggiungerne!!

Documenti analoghi

Eliciting and Analyzing Emergent Misalignment in State-of-the-Art Large Language Models
di: Panpatil, Siddhant, et al.
Pubblicazione: (2025)

AgentRedBench: Dynamic Redteaming and Integration-Aware Defense for LLM Agents over SaaS Integrations
di: Dingeto, Hiskias, et al.
Pubblicazione: (2026)

COMPASS: A Framework for Evaluating Organization-Specific Policy Alignment in LLMs
di: Choi, Dasol, et al.
Pubblicazione: (2026)

PHASOR: Phase-Anchored Universal Action Representations for Humanoid Embodiments
di: Kim, Kihyun, et al.
Pubblicazione: (2026)

REFIND at SemEval-2025 Task 3: Retrieval-Augmented Factuality Hallucination Detection in Large Language Models
di: Lee, DongGeon, et al.
Pubblicazione: (2025)

When Context Flips, Safety Breaks: Diagnosing Brittle Safety in Aligned Language Models
di: Choi, Dasol, et al.
Pubblicazione: (2026)

Everyday Physics in Korean Contexts: A Culturally Grounded Physical Reasoning Benchmark
di: Jeong, Jihae, et al.
Pubblicazione: (2025)

Are Vision-Language Models Safe in the Wild? A Meme-Based Benchmark Study
di: Lee, DongGeon, et al.
Pubblicazione: (2025)

ToDi: Token-wise Distillation via Fine-Grained Divergence Control
di: Jung, Seongryong, et al.
Pubblicazione: (2025)

Jailbreaking on Text-to-Video Models via Scene Splitting Strategy
di: Lee, Wonjun, et al.
Pubblicazione: (2025)

Typed-RAG: Type-Aware Decomposition of Non-Factoid Questions for Retrieval-Augmented Generation
di: Lee, DongGeon, et al.
Pubblicazione: (2025)

When Cars Have Stereotypes: Auditing Demographic Bias in Objects from Text-to-Image Models
di: Choi, Dasol, et al.
Pubblicazione: (2025)

Testing the Limits of Jailbreaking Defenses with the Purple Problem
di: Kim, Taeyoun, et al.
Pubblicazione: (2024)

Infinite-level Fock spaces, crystal bases, and tensor product of extremal weight modules of type $A_{+\infty}$
di: Kwon, Jae-Hoon, et al.
Pubblicazione: (2025)

Theme-Explanation Structure for Table Summarization using Large Language Models: A Case Study on Korean Tabular Data
di: Kwack, TaeYoon, et al.
Pubblicazione: (2025)

Learning While Transmitting: Pilotless Polar Coded Modulation for Short Packet Transmission
di: Choi, Geon, et al.
Pubblicazione: (2026)

Rate-Matching Deep Polar Codes via Polar Coded Extension
di: Choi, Geon, et al.
Pubblicazione: (2025)

Sparsely Pre-transformed Polar Codes for Low-Latency SCL Decoding
di: Choi, Geon, et al.
Pubblicazione: (2024)

Better Safe Than Sorry? Overreaction Problem of Vision Language Models in Visual Emergency Recognition
di: Choi, Dasol, et al.
Pubblicazione: (2025)

MARIOH: Multiplicity-Aware Hypergraph Reconstruction
di: Lee, Kyuhan, et al.
Pubblicazione: (2025)

from Benign import Toxic: Jailbreaking the Language Model via Adversarial Metaphors
di: Yan, Yu, et al.
Pubblicazione: (2025)

Discrimination between ventricular tachycardia and wide‐QRS preexcited tachycardia
di: Jae Hoon Lee
Pubblicazione: (2024)

Benign-to-Toxic Jailbreaking: Inducing Harmful Responses from Harmless Prompts
di: Kim, Hee-Seon, et al.
Pubblicazione: (2025)

ObjexMT: Objective Extraction and Metacognitive Calibration for LLM-as-a-Judge under Multi-Turn Jailbreaks
di: Kim, Hyunjun, et al.
Pubblicazione: (2025)

X-Teaming Evolutionary M2S: Automated Discovery of Multi-turn to Single-turn Jailbreak Templates
di: Kim, Hyunjun, et al.
Pubblicazione: (2025)

Improvement of Glucose Metabolism by Pennogenin 3‐O‐β‐Chacotrioside via Activation of IRS/PI3K/Akt Signaling and Mitochondrial Respiration in Insulin‐Resistant Hepatocytes
di: Jae‐In Lee, et al.
Pubblicazione: (2025)

DeFT-Mamba: Universal Multichannel Sound Separation and Polyphonic Audio Classification
di: Lee, Dongheon, et al.
Pubblicazione: (2024)

Oscillator representations of quantum affine orthosymplectic superalgebras
di: Kwon, Jae-Hoon, et al.
Pubblicazione: (2023)

Accelerating High-Fidelity Waveform Generation via Adversarial Flow Matching Optimization
di: Lee, Sang-Hoon, et al.
Pubblicazione: (2024)

What's Making That Sound Right Now? Video-centric Audio-Visual Localization
di: Choi, Hahyeon, et al.
Pubblicazione: (2025)

Mine-JEPA: In-Domain Self-Supervised Learning for Mine-Like Object Classification in Side-Scan Sonar
di: Kwon, Taeyoun, et al.
Pubblicazione: (2026)

Heuristic Algorithm-based Action Masking Reinforcement Learning (HAAM-RL) with Ensemble Inference Method
di: Choi, Kyuwon, et al.
Pubblicazione: (2024)

When LLMs Go Online: The Emerging Threat of Web-Enabled LLMs
di: Kim, Hanna, et al.
Pubblicazione: (2024)

A New Paradigm Integrating the Concepts of Particle Abrasion and Breakage
di: Tripathi, Priya, et al.
Pubblicazione: (2023)

Hybrid-Vector Retrieval for Visually Rich Documents: Combining Single-Vector Efficiency and Multi-Vector Accuracy
di: Kim, Juyeon, et al.
Pubblicazione: (2025)

DQE-CIR: Distinctive Query Embeddings through Learnable Attribute Weights and Target Relative Negative Sampling in Composed Image Retrieval
di: Park, Geon, et al.
Pubblicazione: (2026)

Self-Guided Target Sound Extraction and Classification Through Universal Sound Separation Model and Multiple Clues
di: Kwon, Younghoo, et al.
Pubblicazione: (2025)

Double-pass rotating z-cut quartz plate as a rapidly variable waveplate
di: Lee, Byungjin, et al.
Pubblicazione: (2025)

El impacto de la variable de género en la migración Honduras-México: el caso de las Hondureñas en Frontera Comalapa
di: Nicanor Madueño Haon
Pubblicazione: (2010)

Distribution-Level Feature Distancing for Machine Unlearning: Towards a Better Trade-off Between Model Utility and Forgetting
di: Choi, Dasol, et al.
Pubblicazione: (2024)