Paper Library

Toward Honest Language Models for Deductive Reasoning

Jiarui Liu, Kaustubh Dhole, Yingheng Wang, Haoyang Wen, Sarah Zhang, Haitao Mao, Gaotang Li, Neeraj Varshney, Jingguo Liu, Xiaoman Pan

2025-11-12

2511.09222v3

Toward Honest Language Models for Deductive Reasoning

Jiarui Liu, Kaustubh Dhole, Yingheng Wang, Haoyang Wen, Sarah Zhang, Haitao Mao, Gaotang Li, Neeraj Varshney, Jingguo Liu, Xiaoman Pan

2025-11-12

2511.09222v2

StyleBreak: Revealing Alignment Vulnerabilities in Large Audio-Language Models via Style-Aware Audio Jailbreak

Hongyi Li, Chengxuan Zhou, Chu Wang, Sicheng Liang, Yanting Chen, Qinlin Xie, Jiawei Ye, Jie Wu

2025-11-12

red teaming

2511.10692v1

iSeal: Encrypted Fingerprinting for Reliable LLM Ownership Verification

Zixun Xiong, Gaoyi Wu, Qingyang Yu, Mingyu Derek Ma, Lingfeng Yao, Miao Pan, Xiaojiang Du, Hao Wang

2025-11-12

2511.08905v2

iSeal: Encrypted Fingerprinting for Reliable LLM Ownership Verification

Zixun Xiong, Gaoyi Wu, Qingyang Yu, Mingyu Derek Ma, Lingfeng Yao, Miao Pan, Xiaojiang Du, Hao Wang

2025-11-12

2511.08905v1

Patching LLM Like Software: A Lightweight Method for Improving Safety Policy in Large Language Models

Huzaifa Arif, Keerthiram Murugesan, Ching-Yun Ko, Pin-Yu Chen, Payel Das, Alex Gittens

2025-11-11

safety

2511.08484v1

SOM Directions are Better than One: Multi-Directional Refusal Suppression in Language Models

Giorgio Piras, Raffaele Mura, Fabio Brau, Luca Oneto, Fabio Roli, Battista Biggio

2025-11-11

2511.08379v2

Why does weak-OOD help? A Further Step Towards Understanding Jailbreaking VLMs

Yuxuan Zhou, Yuzhao Peng, Yang Bai, Kuofeng Gao, Yihao Zhang, Yechao Zhang, Xun Chen, Tao Yu, Tao Dai, Shu-Tao Xia

2025-11-11

red teaming

2511.08367v1

Alignment-Aware Quantization for LLM Safety

Sunghyun Wee, Suyoung Kim, Hyeonjin Kim, Kyomin Hwang, Nojun Kwak

2025-11-11

safety

2511.07842v1

JPRO: Automated Multimodal Jailbreaking via Multi-Agent Collaboration Framework

Yuxuan Zhou, Yang Bai, Kuofeng Gao, Tao Dai, Shu-Tao Xia

2025-11-10

red teaming

2511.07315v1

EduGuardBench: A Holistic Benchmark for Evaluating the Pedagogical Fidelity and Adversarial Safety of LLMs as Simulated Teachers

Yilin Jiang, Mingzi Zhang, Xuanyu Yin, Sheng Jin, Suyu Lu, Zuocan Ying, Zengyi Yu, Xiangjie Kong

2025-11-10

safety

2511.06890v1

Differentiated Directional Intervention A Framework for Evading LLM Safety Alignment

Peng Zhang, Peijie Sun

2025-11-10

safety

2511.06852v2

Differentiated Directional Intervention A Framework for Evading LLM Safety Alignment

Peng Zhang, peijie sun

2025-11-10

red teaming safety

2511.06852v1

SAFENLIDB: A Privacy-Preserving Safety Alignment Framework for LLM-based Natural Language Database Interfaces

Ruiheng Liu, XiaoBing Chen, Jinyu Zhang, Qiongwen Zhang, Yu Zhang, Bailong Yang

2025-11-10

safety

2511.06778v2

SAFENLIDB: A Privacy-Preserving Safety Alignment Framework for LLM-based Natural Language Database Interfaces

Ruiheng Liu, XiaoBing Chen, Jinyu Zhang, Qiongwen Zhang, Yu Zhang, Bailong Yang

2025-11-10

safety

2511.06778v1

November 10 - November 16, 2025

Toward Honest Language Models for Deductive Reasoning

Toward Honest Language Models for Deductive Reasoning

StyleBreak: Revealing Alignment Vulnerabilities in Large Audio-Language Models via Style-Aware Audio Jailbreak

iSeal: Encrypted Fingerprinting for Reliable LLM Ownership Verification

iSeal: Encrypted Fingerprinting for Reliable LLM Ownership Verification

Patching LLM Like Software: A Lightweight Method for Improving Safety Policy in Large Language Models

SOM Directions are Better than One: Multi-Directional Refusal Suppression in Language Models

Why does weak-OOD help? A Further Step Towards Understanding Jailbreaking VLMs

Alignment-Aware Quantization for LLM Safety

JPRO: Automated Multimodal Jailbreaking via Multi-Agent Collaboration Framework

EduGuardBench: A Holistic Benchmark for Evaluating the Pedagogical Fidelity and Adversarial Safety of LLMs as Simulated Teachers

Differentiated Directional Intervention A Framework for Evading LLM Safety Alignment

Differentiated Directional Intervention A Framework for Evading LLM Safety Alignment

SAFENLIDB: A Privacy-Preserving Safety Alignment Framework for LLM-based Natural Language Database Interfaces

SAFENLIDB: A Privacy-Preserving Safety Alignment Framework for LLM-based Natural Language Database Interfaces

November 03 - November 09, 2025

EASE: Practical and Efficient Safety Alignment for Small Language Models

KG-DF: A Black-box Defense Framework against Jailbreak Attacks Based on Knowledge Graphs

Efficient LLM Safety Evaluation through Multi-Agent Debate

RelightMaster: Precise Video Relighting with Multi-plane Light Images

RAG-targeted Adversarial Attack on LLM-based Threat Detection and Mitigation Framework

Injecting Falsehoods: Adversarial Man-in-the-Middle Attacks Undermining Factual Recall in LLMs

Injecting Falsehoods: Adversarial Man-in-the-Middle Attacks Undermining Factual Recall in LLMs

Can LLM Infer Risk Information From MCP Server System Logs?

MCP-RiskCue: Can LLM Infer Risk Information From MCP Server System Logs?