Towards Socially Intelligent Agents with Mental State Transition and Human Utility

Liang, Qing Quan; Zhao, Yang; Liu, Yuan; Lü, Ping; Shi, Wei; Zhou, Yu; Zhu, Song‐Chun

doi:10.48550/arxiv.2103.07011

Cited by 3 publications

(2 citation statements)

References 26 publications

(37 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Theory of mind as a subject, especially in computation (Laird et al, 2017), has gained increased attention in areas including agent-agent reinforcement learning (Rabinowitz et al, 2018), dialogue systems (Qiu et al, 2021), human-computer interaction , agent-agent collaborative dialogue (Roman et al, 2020), and explainable AI (Akula et al, 2021). Worthy of note is the type of mental state recording we employ: specifically, we ask players to record their own mental states during interaction.…”

Section: Related Workmentioning

confidence: 99%

MindCraft: Theory of Mind Modeling for Situated Dialogue in Collaborative Tasks

Bara¹,

CH-Wang²,

Chai³

2021

Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing

View full text Add to dashboard Cite

An ideal integration of autonomous agents in a human world implies that they are able to collaborate on human terms. In particular, theory of mind plays an important role in maintaining common ground during human collaboration and communication. To enable theory of mind modeling in situated interactions, we introduce a fine-grained dataset of collaborative tasks performed by pairs of human subjects in the 3D virtual blocks world of Minecraft. It provides information that captures partners' beliefs of the world and of each other as an interaction unfolds, bringing abundant opportunities to study human collaborative behaviors in situated language communication. As a first step towards our goal of developing embodied AI agents able to infer belief states of collaborative partners in situ, we build and present results on computational models for several theory of mind tasks.

show abstract

Section: Related Workmentioning

confidence: 99%

MindCraft: Theory of Mind Modeling for Situated Dialogue in Collaborative Tasks

Bara¹,

CH-Wang²,

Chai³

2021

Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing

View full text Add to dashboard Cite

show abstract

“…A person's value priority or hierarchy profoundly affects his or her attitudes, beliefs, and traits, making it one core component of personality (Schwartz 2012). In dialogue systems, modeling human values is a critical step towards building socially intelligent chatbots (Qiu et al 2021b). By considering values, we can estimate user behavior and cognitive patterns from their utterances and generate responses that conform to the robot's persona configuration.…”

Section: Introductionmentioning

confidence: 99%

ValueNet: A New Dataset for Human Value Driven Dialogue System

Qiu¹,

Zhao²,

et al. 2022

AAAI

Self Cite

View full text Add to dashboard Cite

Building a socially intelligent agent involves many challenges, one of which is to teach the agent to speak guided by its value like a human. However, value-driven chatbots are still understudied in the area of dialogue systems. Most existing datasets focus on commonsense reasoning or social norm modeling. In this work, we present a new large-scale human value dataset called ValueNet, which contains human attitudes on 21,374 text scenarios. The dataset is organized in ten dimensions that conform to the basic human value theory in intercultural research. We further develop a Transformer-based value regression model on ValueNet to learn the utility distribution. Comprehensive empirical results show that the learned value model could benefit a wide range of dialogue tasks. For example, by teaching a generative agent with reinforcement learning and the rewards from the value model, our method attains state-of-the-art performance on the personalized dialog generation dataset: Persona-Chat. With values as additional features, existing emotion recognition models enable capturing rich human emotions in the context, which further improves the empathetic response generation performance in the EmpatheticDialogues dataset. To the best of our knowledge, ValueNet is the first large-scale text dataset for human value modeling, and we are the first one trying to incorporate a value model into emotionally intelligent dialogue systems. The dataset is available at https://liang-qiu.github.io/ValueNet/.

show abstract

ValueNet: A New Dataset for Human Value Driven Dialogue System

Qiu,

Zhao,

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Building a socially intelligent agent involves many challenges, one of which is to teach the agent to speak guided by its value like a human. However, value-driven chatbots are still understudied in the area of dialogue systems. Most existing datasets focus on commonsense reasoning or social norm modeling. In this work, we present a new large-scale human value dataset called VALUENET, which contains human attitudes on 21,374 text scenarios. The dataset is organized in ten dimensions that conform to the basic human value theory in intercultural research. We further develop a Transformerbased value regression model on VALUENET to learn the utility distribution. Comprehensive empirical results show that the learned value model could benefit a wide range of dialogue tasks. For example, by teaching a generative agent with reinforcement learning and the rewards from the value model, our method attains state-of-the-art performance on the personalized dialog generation dataset: PERSONA-CHAT. With values as additional features, existing emotion recognition models enable capturing rich human emotions in the context, which further improves the empathetic response generation performance in the EMPATHETICDIALOGUES dataset. To the best of our knowledge, VALUENET is the first largescale text dataset for human value modeling, and we are the first one trying to incorporate a value model into emotionally intelligent dialogue systems. The dataset is available at https://liang-qiu.github.io/ValueNet/.

show abstract

Towards Socially Intelligent Agents with Mental State Transition and Human Utility

Cited by 3 publications

References 26 publications

MindCraft: Theory of Mind Modeling for Situated Dialogue in Collaborative Tasks

MindCraft: Theory of Mind Modeling for Situated Dialogue in Collaborative Tasks

ValueNet: A New Dataset for Human Value Driven Dialogue System

ValueNet: A New Dataset for Human Value Driven Dialogue System

Contact Info

Product

Resources

About