Nicholas Meade, Spandana Gella, Devamanyu Hazarika, Prakhar Gupta, Di Jin, Siva Reddy, Yang Liu, Dilek Hakkani-Tur
| Challenge: | Recent work has highlighted safety issues with large neural-based conversational models. |
| Approach: | They propose a retrieval-based approach for reducing bias and toxicity in chatbot responses . they retrieve demonstrations of safe responses to similar dialogue contexts to generate a response . |
| Outcome: | The proposed method reduces bias and toxicity in three chatbot models . it can be used in compliment to existing dialogue safety approaches, such as RLHF. |
Similar Papers
Bot-Adversarial Dialogue for Safe Conversational Agents (2021.naacl-main)
Copied to clipboard
| Challenge: | a new method for evaluating chatbot safety is proposed to mimic human-generated data . a bot-adversarial dialogue model learns undesirable features from this data, a study finds . |
| Approach: | They propose a human-and-model-in-the-loop framework for evaluating toxicity of chatbots . they propose two methods for safe conversational agents by either training on data or ”baking-in” safety to the generative model itself. |
| Outcome: | The proposed methods are safer than existing models while maintaining usability metrics, the authors say . they show that the proposed methods can be used to make safer models with human-model interactions . |
GrounDial: Human-norm Grounded Safe Dialog Response Generation (2024.findings-eacl)
Copied to clipboard
| Challenge: | Recent conversational AI systems generate unsafe responses agreeing to offensive user input or including toxic content. |
| Approach: | They propose a method where response safety is achieved by grounding responses to commonsense social rules without fine-tuning. |
| Outcome: | The proposed approach is quantitatively and qualitatively safer even without additional data or tuning. |
Constructing Highly Inductive Contexts for Dialogue Safety through Controllable Reverse Generation (2022.findings-emnlp)
Copied to clipboard
| Challenge: | Existing methods to detect toxic generation of pretrained language models rely on templates, data extraction, crowdsourcing workers or automatic generation. |
| Approach: | They propose a method to construct adversarial contexts conditioned on a given response . they augment existing dataset BAD+ and construct a new dataset B AD+ . |
| Outcome: | The proposed method can detect toxic or biased content in large pretrained language models. |
On the Safety of Conversational Models: Taxonomy, Dataset, and Benchmark (2022.findings-acl)
Copied to clipboard
Hao Sun, Guangxuan Xu, Jiawen Deng, Jiale Cheng, Chujie Zheng, Hao Zhou, Nanyun Peng, Xiaoyan Zhu, Minlie Huang
| Challenge: | Dialogue safety problems severely limit the real-world deployment of generative conversational models. |
| Approach: | They propose a taxonomy for dialogue safety specifically designed to capture unsafe behaviors in human-bot dialogue settings. |
| Outcome: | The proposed taxonomy captures unsafe behaviors in human-bot dialogue settings with rich context-sensitive unsafe examples. |
ProsocialDialog: A Prosocial Backbone for Conversational Agents (2022.emnlp-main)
Copied to clipboard
Hyunwoo Kim, Youngjae Yu, Liwei Jiang, Ximing Lu, Daniel Khashabi, Gunhee Kim, Yejin Choi, Maarten Sap
| Challenge: | Existing dialogue systems fail to respond properly to potentially unsafe user utterances . existing systems either ignore or passively agree with unsafe content . |
| Approach: | They introduce a dataset to teach conversational agents to respond to problematic content following social norms. |
| Outcome: | The proposed dataset shows that ProsocialDialog generates more socially acceptable dialogues than existing models. |
Just Say No: Analyzing the Stance of Neural Dialogue Generation in Offensive Contexts (2021.emnlp-main)
Copied to clipboard
| Challenge: | despite progress toward data-driven conversational agents, dialogue models still suffer from issues surrounding safety and offensive language. |
| Approach: | They analyze reddit threads and reddits to determine the stance of offensive dialogue models . they find 42% of human responses agree with toxic comments, compared to 13% with safe comments . |
| Outcome: | The proposed model produces 29% fewer offensive replies than the baseline model. |
DialAug: Mixing up Dialogue Contexts in Contrastive Learning for Robust Conversational Modeling (2022.coling-1)
Copied to clipboard
| Challenge: | a conversational system can learn to rank response candidates for a given dialogue context by computing similarity between their vector representations. |
| Approach: | They propose a framework that incorporates augmented dialogue contexts into the learning objective. |
| Outcome: | The proposed framework outperforms existing methods and is more robust to perturbations seen during inference. |
Improving Neural Conversational Models with Entropy-Based Data Filtering (P19-1)
Copied to clipboard
| Challenge: | Current neural network-based conversational models lack diversity and generate boring responses to open-ended utterances. |
| Approach: | They propose an unsupervised method of filtering dialog datasets by removing generic utterances from training data using an entropy-based approach that does not require human supervision. |
| Outcome: | The proposed method improves dialog quality as chatbots learn to output more diverse responses to open-ended utterances. |
Mitigating Gender Bias for Neural Dialogue Generation with Adversarial Learning (2020.emnlp-main)
Copied to clipboard
| Challenge: | Recent research shows that dialogue systems trained on human conversation data are biased and can produce responses that reflect people’s gender prejudice. |
| Approach: | They propose a novel adversarial learning framework Debiased-Chat to train dialogue models free from gender bias while keeping their performance. |
| Outcome: | The proposed framework significantly reduces gender bias in dialogue models while maintaining the response quality. |
SafeConv: Explaining and Correcting Conversational Unsafe Behavior (2023.acl-long)
Copied to clipboard
| Challenge: | Existing datasets do not provide enough annotation to explain unsafe behavior . current chatbots generate toxic and offensive responses, which can be dangerous . |
| Approach: | They construct a dataset called SafeConv that provides comprehensive annotations for chatbots . they compare safe alternatives to rewrite unsafe responses . |
| Outcome: | The proposed model can explain unsafe behavior and detoxify chatbots, the authors show . the proposed model is able to detect unsafe utterances, extract unsafe spans, and convert unsafe responses to safe versions. |