N-GLARE: An Non-Generative Latent Representation-Efficient LLM Safety Evaluator (2026.acl-long)
Copied to clipboard
| Challenge: | Evaluating the safety robustness of LLMs is critical for their deployment. |
| Approach: | They propose to use latent representations to characterize hidden layer dynamics by analyzing the APT of latent models and introducing the JSS metric. |
| Outcome: | The proposed method exploits the APT (Angular-Probabilistic Trajectory) of latent representations and introduces the JSS (Jensen-Shannon Separability) metric. |
Similar Papers
GRE Score: Generative Risk Evaluation for Large Language Models (2026.findings-acl)
Copied to clipboard
| Challenge: | Large language models have revolutionized generative tasks, but concerns about their trustworthiness and vulnerability to adversarial attacks persist. |
| Approach: | They propose an attack-independent evaluation of LLM robustness using conditional generation for synthetic text creation and a method to quantify the model's resilience. |
| Outcome: | The proposed method achieves a consistent ranking of LLM robustness when compared to the attack-based model ranking on TrustLLM (CITATION). |
Automated Progressive Red Teaming (2025.coling-main)
Copied to clipboard
| Challenge: | Automated red teaming (ART) is effective but time-consuming, costly and lacks scalability. |
| Approach: | They propose an automated red teaming framework that generates adversarial prompts to expose LLM vulnerabilities. |
| Outcome: | The proposed framework explores and exploits LLM vulnerabilities through multi-round interactions. |
Probing the Safety Robustness of LLMs in Latent Space (2026.acl-long)
Copied to clipboard
Tianle Gu, Kexin Huang, Zongqi Wang, Yixu Wang, Jie Li, Xin Wang, Yang Yao, Yujiu Yang, Yan Teng, Yingchun Wang
| Challenge: | Despite substantial progress in safety alignment techniques, aligned large language models can still produce unsafe responses under minor internal perturbations. |
| Approach: | They introduce Activation Steering Attack (ASA) and leverage the Negative Log-Likelihood (NLL) as a diagnostic signal to probe the local sensitivity of safety behaviors in latent space. |
| Outcome: | The proposed method is model-agnostic and supervision-free, enabling a general and reproducible diagnostic metric for analyzing safety robustness. |
Do-Not-Answer: Evaluating Safeguards in LLMs (2024.findings-eacl)
Copied to clipboard
| Challenge: | a dataset evaluating harmful capabilities in large language models is available at https://github.com/Libr-AI/do-not-answer. |
| Approach: | They collect an open-source dataset to evaluate the safeguards in large language models . they find that simple BERT-style classifiers can achieve results comparable to GPT-4 . |
| Outcome: | The proposed dataset compares the safety of six popular LLMs to GPT-4 on automatic safety evaluation. |
SafeAgent: Safeguarding LLM Agents via an Automated Risk Simulator (2026.acl-long)
Copied to clipboard
Xueyang Zhou, Weidong Wang, Lin Lu, Jiawen Shi, Guiyao Tie, Xu Yongtian, Lixing Chen, Pan Zhou, Neil Zhenqiang Gong, Lichao Sun
| Challenge: | SafeAgent improves agent safety through fully automated synthetic data generation. |
| Approach: | They propose a framework that improves agent safety through fully automated synthetic data generation. |
| Outcome: | The proposed framework outperforms closed-source models on two safety benchmarks and one real-world task. |
MART: Improving LLM Safety with Multi-round Automatic Red-Teaming (2024.naacl-long)
Copied to clipboard
| Challenge: | Existing red-teaming methods for large language models often discover safety risks without addressing them. |
| Approach: | They propose a multi-round automatic red-teaming method that incorporates both adversarial prompt writing and safe response generation. |
| Outcome: | The proposed method significantly increases red-teaming scalability and the safety of the target LLM. |
When Prompt Optimization Becomes Jailbreaking: Adaptive Red-Teaming of Large Language Models (2026.eacl-srw)
Copied to clipboard
| Challenge: | Existing safety evaluations rely on fixed collections of harmful prompts . such attacks span single-shot prompts, multi-turn interactions, cross-lingual settings . |
| Approach: | They propose to use black-box prompt optimization techniques to search for safety failures . they use GPT-5.1 to optimize for a continuous danger score . |
| Outcome: | The proposed approach reduces effective safeguards for large language models . the average danger score of Qwen 3 8B increases from 0.09 in its baseline setting to 0.79 after optimization. |
Red Teaming Visual Language Models (2024.findings-acl)
Copied to clipboard
| Challenge: | VLMs (Vision-Language Models) can be induced to generate harmful or inaccurate content through specific test cases. |
| Approach: | They propose a red teaming dataset which encompasses 12 subtasks under 4 primary aspects (faithfulness, privacy, safety, fairness) this dataset is the first to benchmark current VLMs in terms of these 4 aspects . |
| Outcome: | The proposed dataset shows that 10 open-source VLMs struggle with red teaming in different degrees and have up to 31% performance gap with GPT-4V. |
TRIDENT: Enhancing Large Language Model Safety with Tri-Dimensional Diversified Red-Teaming Data Synthesis (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) excel in natural language processing tasks but are vulnerable to harmful content and being exploited for malicious purposes. |
| Approach: | They propose a framework to measure the risk coverage of alignment datasets across three dimensions: Lexical Diversity, Malicious Intent, and Jailbreak Tactics. |
| Outcome: | The proposed framework measures risk coverage across Lexical Diversity, Malicious Intent, and Jailbreak Tactics. |
TriPlay-RL: Tri-Role Self-Play Reinforcement Learning for LLM Safety Alignment (2026.acl-long)
Copied to clipboard
Zhewen Tan, Wenhan Yu, Jianfeng Si, Tongxin Liu, Kaiqi Guan, Huiyan Jin, Jiawen Tao, Xiaokun Yuan, Xiangzheng Zhang, Duohe Ma, Tong Yang, Lin Sun
| Challenge: | Existing approaches to safety alignment of large language models rely on costly manual annotations or human review. |
| Approach: | They propose a closed-loop reinforcement learning framework called TriPlay-RL that enables iterative collaboration among three roles with near-zero manual annotation. |
| Outcome: | The proposed framework achieves 20%–50% improvement in adversarial effectiveness while preserving high output diversity while achieving 10%–30% gains in safety performance without degrading general reasoning capability. |