Self-consistency-based approaches, which involve repeatedly sampling multiple outputs and selecting the most consistent one as the final response, prove to be remarkably effective in improving the factual accuracy of large language models. Nonetheless, existing methods usually have strict constraints on the task format, largely limiting their applicability. In this paper, we present Integrative Decoding (ID), to unlock the potential of self-consistency in open-ended generation tasks. ID operates by constructing a set of inputs, each prepended with a previously sampled response, and then processes them concurrently, with the next token being selected by aggregating of all their corresponding predictions at each decoding step. In essence, this simple approach implicitly incorporates self-consistency in the decoding objective. Extensive evaluation shows that ID consistently enhances factuality over a wide range of language models, with substantial improvements on the TruthfulQA (+11.2%), Biographies (+15.4%) and LongFact (+8.5%) benchmarks. The performance gains amplify progressively as the number of sampled responses increases, indicating the potential of ID to scale up with repeated sampling.

Previous research has demonstrated the potential of AI agents to act as companions that can provide constant emotional support for humans. In this paper, we emphasize the necessity of autonomous adaptation in personal AI companionship, an underexplored yet promising direction. Such adaptability is crucial as it can facilitate more tailored interactions with users and allow the agent to evolve in response to users' changing needs. However, imbuing agents with autonomous adaptability presents unique challenges, including identifying optimal adaptations to meet users' expectations and ensuring a smooth transition during the adaptation process. To address them, we devise a hierarchical framework, AutoPal, that enables controllable and authentic adjustments to the agent's persona based on user interactions. A personamatching dataset is constructed to facilitate the learning of optimal persona adaptations. Extensive experiments demonstrate the effectiveness of AutoPal and highlight the importance of autonomous adaptability in AI companionship.

In recent years, there has been a growing interest in exploring dialogues with more complex goals, such as negotiation, persuasion, and emotional support, which go beyond traditional service-focused dialogue systems. Apart from the requirement for much more sophisticated strategic reasoning and communication skills, a significant challenge of these tasks lies in the difficulty of objectively measuring the achievement of their goals in a quantifiable way, making it difficult for existing research to directly optimize the dialogue procedure towards them. In our work, we emphasize the multifaceted nature of complex dialogue goals and argue that it is more feasible to accomplish them by comprehensively considering and jointly promoting their different aspects. To this end, we propose a novel dialogue framework, Cooper, which coordinates multiple specialized agents, each dedicated to a specific dialogue goal aspect separately, to approach the complex objective. Through this divide-and-conquer manner, we make complex dialogue goals more approachable and elicit greater intelligence via the collaboration of individual agents. Experiments on persuasion and emotional support dialogues demonstrate the superiority of our method over a set of competitive baselines. Our codes are available at https://github.com/YiCheng98/Cooper.

With the rise of telemedicine, the task of developing Dialogue Systems for Medical Diagnosis (DSMD) has received much attention in recent years. Different from early researches that needed to rely on extra human resources and expertise to build the system, recent researches focused on how to build DSMD in a data-driven manner. However, the previous data-driven DSMD methods largely overlooked the system interpretability, which is critical for a medical application, and they also suffered from the data sparsity issue at the same time. In this paper, we explore how to bring interpretability to data-driven DSMD. Specifically, we propose a more interpretable decision process to implement the dialogue manager of DSMD by reasonably mimicking real doctors' inquiry logics, and we devise a model with highly transparent components to conduct the inference. Moreover, we collect a new DSMD dataset, which has a much larger scale, more diverse patterns, and is of higher quality than the existing ones. The experiments show that our method obtains 7.7%, 10.0%, 3.0% absolute improvement in diagnosis accuracy respectively on three datasets, demonstrating the effectiveness of its rational decision process and model design. Our codes and the GMD-12 dataset are available at https://github.com/lwgkzl/BR-Agent.

Medical dialogue systems interact with patients to collect symptoms and provide treatment advice. In this task, medical entities (e.g., diseases, symptoms, and medicines) are the most central part of the dialogues. However, existing datasets either do not provide entity annotation or are too small in scale. In this paper, we present MedDG, an entity-centric medical dialogue dataset, where medical entities are annotated with the help of domain experts. It consists of 17,864 Chinese dialogues, 385,951 utterances, and 217,205 entities, at least one magnitude larger than existing entity-annotated datasets. Based on MedDG, we conduct preliminary research on entity-aware medical dialogue generation by implementing several benchmark models. Extensive experiments show that the entity-aware adaptions on the generation models consistently enhance the response quality but there still remains a large space of improvement for future research. The codes and the dataset are released at https://github.com/lwgkzl/MedDG .

Providing Emotional Support (ES) to soothe people in emotional distress is an essential capability in social interactions. Most existing researches on building ES conversation systems only considered single-turn interactions with users, which was over-simplified. In comparison, multi-turn ES conversation systems can provide ES more effectively, but face several new technical challenges, including: (1) how to adopt appropriate support strategies to achieve the long-term dialogue goal of comforting the user's emotion; (2) how to dynamically model the user's state. In this paper, we propose a novel system MultiESC to address these issues. For strategy planning, drawing inspiration from the A* search algorithm, we propose lookahead heuristics to estimate the future user feedback after using particular strategies, which helps to select strategies that can lead to the best long-term effects. For user state modeling, MultiESC focuses on capturing users' subtle emotional expressions and understanding their emotion causes. Extensive experiments show that MultiESC significantly outperforms competitive baselines in both dialogue generation and strategy planning.

Beyond the common difficulties faced in task-oriented dialogue system, medical dialogue has recently attracted increasing attention due to its huge application potential while posing more challenges in reasoning over medical domain knowledge and logic. Existing works resort to neural language models for dialogue embedding and neglect the explicit logical reasoning, leading to poor explainable and generalization ability. In this work, we propose an explainable Heterogeneous Graph Reasoning (HGR) model to unify the relational dialogue context understanding and entity-correlation reasoning into a heterogeneous graph structure. HGR encodes entity context according to the corresponding utterance and deduces next response after fusing the underlying medical knowledge with entity context by attentional graph propagation. To push forward the future research on expert-sensitive task-oriented dialogue system, we first release a large-scale Medical Dialogue Consultant benchmark (MDG-C) with 16 Gastrointestinal diseases for evaluating consultant capability and a Medical Dialogue Diagnosis benchmark (MDG-D) with 6 diseases for measuring diagnosis capability of models, respectively. Extensive experiments on both MDG-C and MDG-D benchmarks demonstrate the superiority of our HGR over state-of-the-art knowledge grounded approaches in general fields of medical dialogue system.

Developing conversational agents to interact with patients and provide primary clinical advice has attracted increasing attention due to its huge application potential, especially in the time of COVID-19 Pandemic. However, the training of end-to-end neural-based medical dialogue system is restricted by an insufficient quantity of medical dialogue corpus. In this work, we make the first attempt to build and release a large-scale high-quality Medical Dialogue dataset related to 12 types of common Gastrointestinal diseases named MedDG, with more than 17K conversations collected from the online health consultation community. Five different categories of entities, including diseases, symptoms, attributes, tests, and medicines, are annotated in each conversation of MedDG as additional labels. To push forward the future research on building expert-sensitive medical dialogue system, we proposes two kinds of medical dialogue tasks based on MedDG dataset. One is the next entity prediction and the other is the doctor response generation. To acquire a clear comprehension on these two medical dialogue tasks, we implement several state-of-the-art benchmarks, as well as design two dialogue models with a further consideration on the predicted entities. Experimental results show that the pre-train language models and other baselines struggle on both tasks with poor performance in our dataset, and the response quality can be enhanced with the help of auxiliary entity information. From human evaluation, the simple retrieval model outperforms several state-of-the-art generative models, indicating that there still remains a large room for improvement on generating medically meaningful responses.