While large neural-based conversational models have become increasingly proficient dialogue agents, recent work has highlighted safety issues with these systems. For example, these systems can be goaded into generating toxic content, which often perpetuates social biases or stereotypes. We investigate a retrieval-based method for reducing bias and toxicity in responses from chatbots. It uses in-context learning to steer a model towards safer generations. Concretely, to generate a response to an unsafe dialogue context, we retrieve demonstrations of safe responses to similar dialogue contexts. We find our method performs competitively with strong baselines without requiring training. For instance, using automatic evaluation, we find our best fine-tuned baseline only generates safe responses to unsafe dialogue contexts from DiaSafety 4.04% more than our approach. Finally, we also propose a re-ranking procedure which can further improve response safeness.
The Rise of Darkness: Safety-Utility Trade-Offs in Role-Playing Dialogue Agents
Institute of Computing and Intelligence, Harbin Institute of Technology(哈尔滨工业大学计算与智能学院);Baidu Inc.(百度公司);College of Intelligence and Computing, Tianjin University(天津大学智能与计算学院)
CoSafe: Evaluating Large Language Model Safety in Multi-Turn Dialogue Coreference
The Hong Kong Polytechnic University(香港理工大学);Huawei Noah’s Ark Lab(华为诺亚方舟实验室);Research Centre for Data Science & Artificial Intelligence(数据科学与人工智能研究中心)
SafeDialBench: A Fine-Grained Safety Evaluation Benchmark for Large Language Models in Multi-Turn Dialogues with Diverse Jailbreak Attacks
National Key Laboratory for Novel Software Technology, Nanjing University(南京大学新型软件技术国家重点实验室);University of Liverpool(利物浦大学);City University of Hong Kong(香港城市大学);School of Intelligence Science and Technology, Nanjing University(南京大学智能科学与技术学院);China Mobile Research Institute(中国移动研究院);China Mobile (Suzhou) Software Technology Co., Ltd(中国移动苏州软件技术有限公司)
MTMCS-Bench: Evaluating Contextual Safety of Multimodal Large Language Models in Multi-Turn Dialogues
University of Notre Dame(诺丁汉大学);University of California, Los Angeles(加州大学洛杉矶分校);Georgia Institute of Technology(佐治亚理工学院);University of Montreal(蒙特利尔大学)
Persistent Personas? Role-Playing, Instruction Following, and Safety in Extended Interactions
University of Vienna, Faculty of Computer Science(维也纳大学计算机科学系);Doctoral School Computer Science, University of Vienna(维也纳大学计算机科学博士学院);Center for Information and Language Processing, LMU Munich(慕尼黑大学信息与语言处理中心);Munich Center for Machine Learning(慕尼黑机器学习中心);University of Vienna, Faculty of Philological and Cultural Studies(维也纳大学文学与文化研究系)
在本文中,我们介绍了PanGu-Bot,一个基于大型预训练语言模型(PLM)PANGU-alpha(Zeng et al.,2021年)。与其他从零开始在大量对话数据上训练的预训练对话模型不同,我们的目标是通过继承PLM中有价值的语言能力和知识,以相对较少的数据和计算成本构建强大的对话模型。为此,我们从大型PLM PANGU-alpha中训练PanGu-Bot,该系统已被证明在各种中文自然语言任务上表现良好。我们调查PanGu-Bot产生的响应的不同方面,包括响应质量,知识和安全性。我们表明PanGu-Bot优于最先进的中文对话系统(CDIALGPT(Wang et al.,2020)、EVA(Zhou等人,2021)、EVA2.0(Gu等人,2022))w.r.t.以上三个方面。我们还证明,PanGu-Bot可以很容易地部署,以产生情绪反应,而无需进一步培训。通过实证分析,我们也指出盘古机器人的响应质量、知识正确性和安全性还远远不够完善,要构建可靠、智能的对话系统还需要进一步探索。我们的模型和代码将很快在https://github.com/huawei-noah/Pretrained-Language-Model/tree/master/PanGu-Bot上提供。
Learning an Effective Context-Response Matching Model with Self-Supervised Tasks for Retrieval-based Dialogues
Wangxuan Institute of Computer Technology, Peking University(北京大学王选计算机研究所);Microsoft Corporation(微软公司);Center for Data Science, Peking University(北京大学数据科学研究中心)
ACR: Adaptive Context Refactoring via Context Refactoring Operators for Multi-Turn Dialogue
Zhejiang Normal University(浙江师范大学);Alibaba Group(阿里巴巴集团);Hong Kong University of Science and Technology(香港科学与技术大学);Southeast University(东南大学);Soochow University(苏州大学)
Modeling the One-to-Many Property in Open-Domain Dialogue with LLMs
School of Electrical and Electronic Engineering, Nanyang Technological University(南洋理工大学电子与电气工程学院);Department of Electrical and Electronic Engineering, The Hong Kong Polytechnic University(香港理工大学电子与电气工程系)
Improving Dialogue State Tracking through Combinatorial Search for In-Context Examples
Graduate School of Data Science, Seoul National University(数据科学研究生院,首尔国立大学);Department of Computer Science and Engineering, Seoul National University(计算机科学与工程系,首尔国立大学)
Build it Break it Fix it for Dialogue Safety: Robustness from Adversarial Human Attack
Facebook AI Research(脸书人工智能研究院);Virginia Tech(弗吉尼亚理工大学)
在对话的上下文中检测冒犯性语言已经成为自然语言处理的越来越重要的应用。公共论坛中巨魔的检测(Galán-García等人,2016),以及在公共领域部署聊天机器人(Wolf et al.,2017年)是两个例子,表明有必要防范人类的对抗性攻击行为。在这项工作中,我们开发了一个训练方案,通过迭代构建它,打破它,修复它,人类和模型在循环中的策略,使模型对这种人类攻击变得鲁棒。在详细的实验中,我们证明了这种方法比以前的系统更加强大。此外,我们表明,在对话中使用的攻击性语言严重依赖于对话上下文,不能被视为一个单一的句子攻击检测任务,在大多数以前的工作。我们新收集的任务和方法将开源并公开。
Towards Multi-modal Multi-turn Safety: From Agentic Interaction to Strategic Alignment
Hong Kong University of Science and Technology(香港科技大学);Zhongshan School of Medicine, SUN YAT-SEN UNIVERSITY(中山医学院,孙中山大学);University of Edinburgh(爱丁堡大学);University of Washington(华盛顿大学)