EmpathyAgent: Can Embodied Agents Conduct Empathetic Actions?

2025-03-19

AI Summary

This work addresses the critical gap in evaluating empathic capabilities of embodied agents by introducing EmpathyAgent—the first systematic benchmark for embodied empathy assessment. It comprises 10,000 multimodal samples and a structured empathy task taxonomy covering cognitive, affective, and behavioral empathy challenges. Methodologically, we formally define and quantify empathic *action capability* in embodied agents, develop a process-oriented, task-specific evaluation suite, and perform instruction-tuning on Llama3-8B to enhance empathic behavior generation. Experimental results reveal severe deficiencies in empathic action performance among mainstream models; instruction-tuned variants achieve a substantial +32.7% improvement in empathic action accuracy. All code, data, and evaluation protocols are fully open-sourced, establishing the first reproducible, scalable, and standardized foundation for research on embodied empathy.

Technology Category

• Intelligent Robots: Embodied AI
• Humans and AI: Emotional Intelligence
• Multiagent Systems: Agent-Based Simulation and Emergent Behavior

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Search and Retrieval-Augmented AI: Agentic search
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

Empathy is fundamental to human interactions, yet it remains unclear whether embodied agents can provide human-like empathetic support. Existing works have studied agents' tasks solving and social interactions abilities, but whether agents can understand empathetic needs and conduct empathetic behaviors remains overlooked. To address this, we introduce EmpathyAgent, the first benchmark to evaluate and enhance agents' empathetic actions across diverse scenarios. EmpathyAgent contains 10,000 multimodal samples with corresponding empathetic task plans and three different challenges. To systematically evaluate the agents' empathetic actions, we propose an empathy-specific evaluation suite that evaluates the agents' empathy process. We benchmark current models and found that exhibiting empathetic actions remains a significant challenge. Meanwhile, we train Llama3-8B using EmpathyAgent and find it can potentially enhance empathetic behavior. By establishing a standard benchmark for evaluating empathetic actions, we hope to advance research in empathetic embodied agents. Our code and data are publicly available at https://github.com/xinyan-cxy/EmpathyAgent.

Problem

• Can embodied agents perform human-like empathetic actions?
• Lack of benchmarks for evaluating agents' empathetic behaviors
• Enhancing empathetic behavior in AI using multimodal datasets

Innovation

• Introduces EmpathyAgent benchmark for empathetic agents
• Proposes empathy-specific evaluation suite for agents
• Trains Llama3-8B to enhance empathetic behavior