LongSafetyBench: Long-Context LLMs Struggle with Safety Issues

• URL: http://arxiv.org/abs/2411.06899v1
• Date: Mon, 11 Nov 2024 11:57:37 GMT
• Title: LongSafetyBench: Long-Context LLMs Struggle with Safety Issues
• Authors: Mianqiu Huang, Xiaoran Liu, Shaojun Zhou, Mozhi Zhang, Chenkun Tan, Pengyu Wang, Qipeng Guo, Zhe Xu, Linyang Li, Zhikai Lei, Linlin Li, Qun Liu, Yaqian Zhou, Xipeng Qiu, Xuanjing Huang,
• Abstract summary: Long-context language models can exhibit safety concerns. The proportion of safe responses from most mainstream long-context models is below 50%. LongSafetyBench can serve as a valuable benchmark for evaluating the safety capabilities of long-context language models.
• Score: 83.75506157823517
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: With the development of large language models (LLMs), the sequence length of these models continues to increase, drawing significant attention to long-context language models. However, the evaluation of these models has been primarily limited to their capabilities, with a lack of research focusing on their safety. Existing work, such as ManyShotJailbreak, has to some extent demonstrated that long-context language models can exhibit safety concerns. However, the methods used are limited and lack comprehensiveness. In response, we introduce \textbf{LongSafetyBench}, the first benchmark designed to objectively and comprehensively evaluate the safety of long-context models. LongSafetyBench consists of 10 task categories, with an average length of 41,889 words. After testing eight long-context language models on LongSafetyBench, we found that existing models generally exhibit insufficient safety capabilities. The proportion of safe responses from most mainstream long-context LLMs is below 50\%. Moreover, models' safety performance in long-context scenarios does not always align with that in short-context scenarios. Further investigation revealed that long-context models tend to overlook harmful content within lengthy texts. We also proposed a simple yet effective solution, allowing open-source models to achieve performance comparable to that of top-tier closed-source models. We believe that LongSafetyBench can serve as a valuable benchmark for evaluating the safety capabilities of long-context language models. We hope that our work will encourage the broader community to pay attention to the safety of long-context models and contribute to the development of solutions to improve the safety of long-context LLMs.

Related papers

• Scaling Instruction-Tuned LLMs to Million-Token Contexts via Hierarchical Synthetic Data Generation [15.975325252309554]
  We introduce a novel post-training synthetic data generation strategy designed to efficiently extend the context window of Large Language Models. Our approach scalably extends to arbitrarily long context lengths, unconstrained by the length of available real-world data. We demonstrate that our model, with a context length of up to 1M tokens, performs well on the RULER benchmark and InfiniteBench.
  arXiv  Detail & Related papers  (2025-04-17T04:46:57Z)
• InfiniteICL: Breaking the Limit of Context Window Size via Long Short-term Memory Transformation [57.310236384112834]
  In-context learning (ICL) is critical for large language models (LLMs) but its effectiveness is constrained by finite context windows. We introduce InfiniteICL, a framework that parallels context and parameters in LLMs with short- and long-term memory. We demonstrate that our method reduces context length by 90% while achieving 103% average performance of full-context prompting.
  arXiv  Detail & Related papers  (2025-04-02T13:15:44Z)
• LongSafety: Evaluating Long-Context Safety of Large Language Models [95.2469116388522]
  LongSafety is the first benchmark designed to evaluate safety in open-ended long-context tasks. Our evaluation reveals significant safety vulnerabilities, with most models achieving safety rates below 55%. Our findings emphasize the unique challenges and urgency of improving long-context safety.
  arXiv  Detail & Related papers  (2025-02-24T08:54:39Z)
• LongPO: Long Context Self-Evolution of Large Language Models through Short-to-Long Preference Optimization [49.37607974207405]
  LongPO harnesses short-to-long preference data to transfer short-context capabilities to long-context tasks. LongPO fully retains short-context performance and largely outperforms naive SFT and DPO in both long- and short-context tasks.
  arXiv  Detail & Related papers  (2025-02-19T17:59:03Z)
• LIFT: Improving Long Context Understanding Through Long Input Fine-Tuning [35.31849814789343]
  This paper introduces Long Input Fine-Tuning (LIFT) for long context modeling. LIFT enables efficient processing of lengthy inputs without the computational burden of offline long-context adaptation. The framework is further enhanced by integrating in-context learning and pre-LIFT supervised fine-tuning.
  arXiv  Detail & Related papers  (2024-12-18T09:04:55Z)
• How Effective Is Self-Consistency for Long-Context Problems? [18.633918831942434]
  Self-consistency (SC) has been demonstrated to enhance the performance of large language models (LLMs) This study examines the role of SC in long-context scenarios, where LLMs often struggle with position bias.
  arXiv  Detail & Related papers  (2024-11-02T01:52:42Z)
• LongReward: Improving Long-context Large Language Models with AI Feedback [54.3321542678909]
  LongReward is a novel method that provides rewards for long-context model responses from four human-valued dimensions. Our experiments indicate that LongReward not only significantly improves models' long-context performance but also enhances their ability to follow short instructions.
  arXiv  Detail & Related papers  (2024-10-28T17:50:42Z)
• Multimodal Situational Safety [73.63981779844916]
  We present the first evaluation and analysis of a novel safety challenge termed Multimodal Situational Safety. For an MLLM to respond safely, whether through language or action, it often needs to assess the safety implications of a language query within its corresponding visual context. We develop the Multimodal Situational Safety benchmark (MSSBench) to assess the situational safety performance of current MLLMs.
  arXiv  Detail & Related papers  (2024-10-08T16:16:07Z)
• Leave No Document Behind: Benchmarking Long-Context LLMs with Extended Multi-Doc QA [71.04146366608904]
  Long-context modeling capabilities have garnered widespread attention, leading to the emergence of Large Language Models (LLMs) with ultra-context windows. We propose a novel long-context benchmark, Loong, aligning with realistic scenarios through extended multi-document question answering (QA) Loong introduces four types of tasks with a range of context lengths: Spotlight Locating, Comparison, Clustering, and Chain of Reasoning.
  arXiv  Detail & Related papers  (2024-06-25T09:42:56Z)
• Exploring Advanced Methodologies in Security Evaluation for LLMs [16.753146059652877]
  Large Language Models (LLMs) represent an advanced evolution of earlier, simpler language models. They boast enhanced abilities to handle complex language patterns and generate coherent text, images, audios, and videos. Rapid expansion of LLMs has raised security and ethical concerns within the academic community.
  arXiv  Detail & Related papers  (2024-02-28T01:32:58Z)
• Training With "Paraphrasing the Original Text" Improves Long-Context Performance [19.48556587305737]
  Large Language Models (LLMs) continue to evolve, more are being designed to handle long-context inputs. We propose a novel approach to design training data for long-context tasks, aiming at augmenting LLMs' proficiency in extracting key information from long context. Experimenting on LongBench and NaturalQuestions Multi-document-QA dataset with models of Llama and Qwen series, our method achieves an improvement of up to 8.48% and 4.48% in average scores.
  arXiv  Detail & Related papers  (2023-12-18T13:40:16Z)
• BAMBOO: A Comprehensive Benchmark for Evaluating Long Text Modeling Capacities of Large Language Models [141.21603469555225]
  Large language models (LLMs) have achieved dramatic proficiency over NLP tasks with normal length. We propose BAMBOO, a multi-task long context benchmark. It consists of 10 datasets from 5 different long text understanding tasks.
  arXiv  Detail & Related papers  (2023-09-23T11:36:15Z)
• SafetyBench: Evaluating the Safety of Large Language Models [54.878612385780805]
  SafetyBench is a comprehensive benchmark for evaluating the safety of Large Language Models (LLMs) It comprises 11,435 diverse multiple choice questions spanning across 7 distinct categories of safety concerns. Our tests over 25 popular Chinese and English LLMs in both zero-shot and few-shot settings reveal a substantial performance advantage for GPT-4 over its counterparts.
  arXiv  Detail & Related papers  (2023-09-13T15:56:50Z)
• LongBench: A Bilingual, Multitask Benchmark for Long Context Understanding [58.20031627237889]
  LongBench is the first bilingual, multi-task benchmark for long context understanding. It comprises 21 datasets across 6 task categories in both English and Chinese, with an average length of 6,711 words (English) and 13,386 characters (Chinese)
  arXiv  Detail & Related papers  (2023-08-28T11:53:40Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.