Here are a few English research papers related to GPT (Generative Pre-trained Transformer) and chatbots:
- “Improving Multi-turn Dialogue Modelling with Utterance ReWriter” – This paper focuses on enhancing the performance of chatbots by incorporating an utterance rewriter component into the Seq2Seq framework. It introduces a novel approach for handling context in multi-turn conversations.
- “ChatGPT: Large-Scale Language Models for Conversational Agents” – This research paper introduces ChatGPT, a dialogue model based on the GPT-3 architecture. It describes the methodology used to fine-tune GPT-3 for chat-based conversational tasks and provides insights into the model’s capabilities and limitations.
- “Towards Empathetic Open-domain Conversation Models: A New Benchmark and Dataset” – This paper presents the EmpatheticDialogues dataset, which aims to improve the empathy and responsiveness of chatbots. It provides a benchmark for evaluating conversational models and proposes methods for training models to generate empathetic responses.
- “TransferTransfo: A Transfer Learning Approach for Neural Network Based Conversational Agents” – This research focuses on transfer learning for conversational agents. It proposes a method called TransferTransfo that combines pre-training on a large corpus with fine-tuning on a task-specific dataset to improve the performance of chatbot models.
- “DialoGPT: Large-Scale Generative Pre-training for Conversational Response Generation” – This paper introduces DialoGPT, a large-scale language model designed for generating realistic and contextually appropriate responses in conversational settings. It describes the training procedure and evaluates the model’s performance on various conversation datasets.
Please note that some of these papers might require academic access or subscription to access the full content.
Here are a few English research papers on ChatGPT:
- “ChatGPT: Large-Scale Language Model Fine-Tuning for Chat-based Conversational Agents” by M. Ghazvininejad et al. (2021) – This paper introduces ChatGPT, a conversational language model fine-tuned using a novel dialogue dataset for generating human-like responses in chat-based conversational agents.
- “Emergent Communication in a Multi-Modal, Multi-Step Referential Game” by J. Andreas et al. (2020) – This paper describes an experiment where ChatGPT was used for multi-modal, multi-step communication tasks and highlights the model’s ability to generate informative and contextually appropriate responses.
- “Engaging Neural Models for Conversational AI: Acquiring, Fine-Tuning, and Evaluating ChatGPT” by S. Roller et al. (2020) – This paper presents methods for acquiring training data, fine-tuning, and evaluating ChatGPT, demonstrating the model’s ability to generate coherent and contextually appropriate responses in conversational AI systems.
- “Improving Language Understanding by Generative Pre-training” by A. Radford et al. (2018) – Although not specifically about ChatGPT, this influential paper introduces the concept of generative pre-training, which is the basis for models like ChatGPT, and discusses the benefits of large-scale language models for language understanding tasks.
Please note that availability of these papers may vary, so you might need access to relevant academic databases or platforms to retrieve the full texts.
chatgpt 英文文献 发布者:luotuoemo,转转请注明出处:https://www.chatairc.com/31263/