Kardia-R1: Unleashing LLMs to Reason toward Understanding and Empathy for Emotional Support via Rubric-as-Judge Reinforcement Learning

TL;DR

This paper introduces Kardia-R1, a novel framework that enhances large language models' empathetic reasoning through a large-scale, user-grounded benchmark and rubric-guided reinforcement learning, improving emotional understanding and persona consistency.

Contribution

It presents KardiaBench, a large, psychologically plausible dataset, and Kardia-R1, a new training method using explainable rubric rewards for empathetic reasoning in LLMs.

Findings

Kardia-R1 outperforms existing methods in emotion accuracy and empathy.

The dataset ensures psychological plausibility and persona consistency.

Rubric-guided RL improves interpretability and alignment with human empathy standards.

Abstract

As web platforms evolve towards greater personalization and emotional complexity, conversational agents must transcend superficial empathy to demonstrate identity-aware emotional reasoning. However, existing systems face two limitations: (1) reliance on situation-centric datasets lacking persistent user identity, which hampers the capture of personalized affective nuances; and (2) dependence on opaque, coarse reward signals that hinder development of verifiable empathetic reasoning. To address these gaps, we introduce KardiaBench, a large-scale user-grounded benchmark comprising 178,080 QA pairs across 22,080 multi-turn conversations anchored to 671 real-world profiles. The dataset is constructed via a model-in-the-loop pipeline with iterative rubric-guided refinement to ensure psychological plausibility and persona consistency. This progressive empathy pipeline that integrates user comprehension, contextual reasoning, and emotion perception into conversations, followed by iterative critique and rubric-based refinement to ensure psychological plausibility, emotional fidelity, and persona consistency. Building on this, we propose Kardia-R1, a framework that trains models for interpretable, stepwise empathetic cognition. Kardia-R1 leverages Rubric-as-Judge Empathetic Reinforcement Learning (Rubric-ERL), a GRPO-based method that uses explainable, human-aligned rubric rewards to tightly couple user understanding, emotional inference, and supportive response generation. Extensive experiments across four LLM backbones demonstrate that Kardia-R1 consistently outperforms othet methods in emotion accuracy, empathy, relevance, persona consistency, and safety. Our dataset and model will be released at https://github.com/JhCircle/Kardia-R1.