Value Imprint: A Technique for Auditing the Human Values Embedded in RLHF Datasets

LLMs are increasingly fine-tuned using RLHF datasets to align them with human preferences and values. However, very limited research has investigated which specific human values are operationalized through these datasets. In this paper, we introduce Value Imprint, a framework for auditing and classifying the human values embedded within RLHF datasets. To investigate the viability of this framework, we conducted three case study experiments by auditing the Anthropic/hh-rlhf, OpenAI WebGPT Comparisons, and Alpaca GPT-4-LLM datasets to examine the human values embedded within them. Our analysis involved a two-phase process. During the first phase, we developed a taxonomy of human values through an integrated review of prior works from philosophy, axiology, and ethics. Then, we applied this taxonomy to annotate 6,501 RLHF preferences. During the second phase, we employed the labels generated from the annotation as ground truth data for training a transformer-based machine learning model to audit and classify the three RLHF datasets. Through this approach, we discovered that information-utility values, including Wisdom/Knowledge and Information Seeking, were the most dominant human values within all three RLHF datasets. In contrast, prosocial and democratic values, including Well-being, Justice, and Human/Animal Rights, were the least represented human values. These findings have significant implications for developing language models that align with societal values and norms. We contribute our datasets to support further research in this area.

Investigating the Impact of Trust in Multi-Human Multi-Robot Task Allocation

Trust is essential in human-robot collaboration. Even more so in multi-human multi-robot teams where trust is vital to ensure teaming cohesion in complex operational environments. Yet, at the moment, trust is rarely considered a factor during task allocation and reallocation in algorithms used in multi-human, multi-robot collaboration contexts. Prior work on trust in single-human-robot interaction has identified that including trust as a parameter in human-robot interaction significantly improves both performance outcomes and human experience with robotic systems. However, very little research has explored the impact of trust in multi-human multi-robot collaboration, specifically in the context of task allocation. In this paper, we introduce a new trust model, the Expectation Comparison Trust (ECT) model, and employ it with three trust models from prior work and a baseline no-trust model to investigate the impact of trust on task allocation outcomes in multi-human multi-robot collaboration. Our experiment involved different team configurations, including 2 humans, 2 robots, 5 humans, 5 robots, and 10 humans, 10 robots. Results showed that using trust-based models generally led to better task allocation outcomes in larger teams (10 humans and 10 robots) than in smaller teams. We discuss the implications of our findings and provide recommendations for future work on integrating trust as a variable for task allocation in multi-human, multi-robot collaboration.

Adaptive Task Allocation in Multi-Human Multi-Robot Teams under Team Heterogeneity and Dynamic Information Uncertainty

Task allocation in multi-human multi-robot (MH-MR) teams presents significant challenges due to the inherent heterogeneity of team members, the dynamics of task execution, and the information uncertainty of operational states. Existing approaches often fail to address these challenges simultaneously, resulting in suboptimal performance. To tackle this, we propose ATA-HRL, an adaptive task allocation framework using hierarchical reinforcement learning (HRL), which incorporates initial task allocation (ITA) that leverages team heterogeneity and conditional task reallocation in response to dynamic operational states. Additionally, we introduce an auxiliary state representation learning task to manage information uncertainty and enhance task execution. Through an extensive case study in large-scale environmental monitoring tasks, we demonstrate the benefits of our approach.

PrefCLM: Enhancing Preference-based Reinforcement Learning with Crowdsourced Large Language Models

Preference-based reinforcement learning (PbRL) is emerging as a promising approach to teaching robots through human comparative feedback, sidestepping the need for complex reward engineering. However, the substantial volume of feedback required in existing PbRL methods often lead to reliance on synthetic feedback generated by scripted teachers. This approach necessitates intricate reward engineering again and struggles to adapt to the nuanced preferences particular to human-robot interaction (HRI) scenarios, where users may have unique expectations toward the same task. To address these challenges, we introduce PrefCLM, a novel framework that utilizes crowdsourced large language models (LLMs) as simulated teachers in PbRL. We utilize Dempster-Shafer Theory to fuse individual preferences from multiple LLM agents at the score level, efficiently leveraging their diversity and collective intelligence. We also introduce a human-in-the-loop pipeline that facilitates collective refinements based on user interactive feedback. Experimental results across various general RL tasks show that PrefCLM achieves competitive performance compared to traditional scripted teachers and excels in facilitating more more natural and efficient behaviors. A real-world user study (N=10) further demonstrates its capability to tailor robot behaviors to individual user preferences, significantly enhancing user satisfaction in HRI scenarios.

Robot Patrol: Using Crowdsourcing and Robotic Systems to Provide Indoor Navigation Guidance to The Visually Impaired

Indoor navigation is a challenging activity for persons with disabilities, particularly, for those with low vision and visual impairment. Researchers have explored numerous solutions to resolve these challenges; however, several issues remain unsolved, particularly around providing dynamic and contextual information about potential obstacles in indoor environments. In this paper, we developed Robot Patrol, an integrated system that employs a combination of crowdsourcing, computer vision, and robotic frameworks to provide contextual information to the visually impaired to empower them to navigate indoor spaces safely. In particular, the system is designed to provide information to the visually impaired about 1) potential obstacles on the route to their indoor destination, 2) information about indoor events on their route which they may wish to avoid or attend, and 3) any other contextual information that might support them to navigate to their indoor destinations safely and effectively. Findings from the Wizard of Oz experiment of our demo system provide insights into the benefits and limitations of the system. We provide a concise discussion on the implications of our findings.