---
title: 'Empathy-R1: Structured Empathy for Counseling'
url: https://www.emergentmind.com/topics/empathy-r1
type: topic
---

# Empathy-R1: Structured Empathy for Counseling

Searching arXiv for the primary paper and closely related work on empathetic reasoning and RL in dialogue.
Empathy-R1 is a framework for long-form mental health support that integrates a Chain-of-Empathy (CoE) reasoning process with reinforcement learning to improve response quality on Long Counseling Texts (LCTs), particularly in a Chinese context [2509.14851]. It is designed to address a specific limitation of existing large language models: replies may be semantically fluent yet still lack the structured reasoning required for genuine psychological support. The framework combines a psychologically grounded multi-step reasoning scaffold, a new large-scale Chinese dataset named Empathy-QA, and a two-stage training procedure in which Supervised Fine-Tuning instills reasoning structure and reinforcement learning refines therapeutic relevance and contextual appropriateness [2509.14851]. Within the recent literature on empathetic dialogue, Empathy-R1 belongs to a broader shift from surface-level affective mirroring toward explicit reasoning, structured empathy evaluation, and reward-based alignment [2408.02976] [2512.01282] [2601.18281].

## 1. Definition and scope

Empathy-R1 is defined by two core commitments: interpretable reasoning and reinforcement learning for long-form mental health response generation [2509.14851]. The framework is motivated by the observation that, in mental health support, fluency alone is insufficient. Effective assistance requires psychological depth, genuine empathy, and a reasoning process that can account for the help-seeker’s emotions, causes, and intentions before formulating a supportive reply [2509.14851].

The system is explicitly targeted at Long Counseling Texts, a setting in which user inputs are extended, context-rich, and often psychologically complex [2509.14851]. This distinguishes it from many earlier empathetic response generation systems developed for shorter open-domain dialogue turns. A plausible implication is that Empathy-R1 treats empathy not as a single stylistic attribute of the final answer, but as a structured inferential process that precedes response generation.

In the literature, this design sits alongside several related but distinct approaches. EmpRL aligns empathy levels between generated and target responses through a reinforcement learning reward defined over emotional reaction, interpretation, and exploration [2408.02976]. Kardia-R1 trains models for interpretable, stepwise empathetic cognition using rubric-based reinforcement learning over relevance, fluency, empathy, persona consistency, and safety [2512.01282]. ReEmpathy interleaves spoken response generation with reflective empathy-related reasoning in end-to-end spoken dialogue [2601.18281]. Empathy-R1 differs in centering long-form Chinese mental health support and in organizing reasoning through the specific CoE scaffold [2509.14851].

## 2. Chain-of-Empathy reasoning paradigm

The central mechanism in Empathy-R1 is the Chain-of-Empathy reasoning paradigm, a structured framework inspired by cognitive-behavioral therapy and professional counseling practices [2509.14851]. Rather than leaving empathetic behavior implicit in the model’s hidden states, CoE makes the analytic pathway explicit and interpretable.

The reasoning structure is four-layered:

| Layer | Function |
|---|---|
| L1 | Emotions and Context |
| L2 | Causes and Beliefs |
| L3 | Intent Analysis |
| L4 | Response Strategy |

At **L1: Emotions and Context**, the model recognizes the user’s core feelings and situational context, grounded in appraisal theory [2509.14851]. At **L2: Causes and Beliefs**, it analyzes underlying causes such as cognitive biases or beliefs that generate the emotional state [2509.14851]. At **L3: Intent Analysis**, it discerns the user’s primary goal, including validation, advice, or understanding [2509.14851]. At **L4: Response Strategy**, it synthesizes a therapeutic response by combining the preceding analyses and drawing on techniques such as active listening, Rogers, and the NURSE model [2509.14851].

Implementation-wise, the model is prompted to reason step by step through these layers and to produce both an `empathy_think` rationale and a user-facing answer [2509.14851]. The paper states that this is enforced through explicit system prompts and output format constraints. This places Empathy-R1 in a lineage of systems that make empathy operationally inspectable. IMAGINE, for example, explicitly models emotion cause extraction, knowledge extension, and communication mechanisms such as emotional reaction, interpretation, and exploration [2212.02000]. ReEmpathy similarly uses explicit reflection traces during spoken response generation [2601.18281]. Empathy-R1’s specific contribution is to adapt such explicit reasoning to long counseling texts and to frame it as a therapeutic scaffold [2509.14851].

## 3. Empathy-QA dataset

Empathy-R1 is supported by a new Chinese dataset, Empathy-QA, introduced as a large-scale resource for model development and evaluation in long-form mental health support [2509.14851]. The paper positions this dataset as a response to the limitations of prior resources such as PsyQA, which are described as outdated and limited in scope [2509.14851].

Empathy-QA contains **40,959 user questions** and **168,470 long-form responses** [2509.14851]. It covers **contemporary topics**, including **AI-induced professional anxiety**, and is filtered according to quality, length, and ethical standards [2509.14851]. Each instance includes a question title, description, topic label, and multiple human responses [2509.14851]. The dataset serves both as a pretraining corpus and as a benchmark [2509.14851].

This emphasis on domain-specific, large-scale, and culturally situated data is consistent with a broader trend in empathetic dialogue research. KardiaBench, used by Kardia-R1, grounds empathy in persistent user identity and evolving conversation through **178,080 QA pairs across 22,080 multi-turn conversations anchored to 671 real-world profiles** [2512.01282]. Psyche-R1 similarly uses **73,000+ high-quality empathetic dialogues** alongside reasoning-oriented psychological QA data for Chinese psychological modeling [2508.10848]. Empathy-R1’s dataset contribution is narrower in task scope but specialized for long counseling texts and Chinese mental health support [2509.14851].

A plausible implication is that Empathy-QA does not merely expand scale; it also shifts the empirical target from short empathetic replies to sustained counseling-style generation, where contextual continuity and therapeutic appropriateness are central.

## 4. Training procedure and reward design

Empathy-R1 uses a two-stage training process consisting of Supervised Fine-Tuning followed by reinforcement learning with Group Relative Policy Optimization (GRPO) [2509.14851].

During **Stage 1: Supervised Fine-Tuning**, the goal is to teach the model the explicit CoE reasoning structure and output format [2509.14851]. The training set includes a high-quality subset of **200 curated examples** generated via Deepseek-R1 and reviewed by psychology professionals [2509.14851]. Each sample contains the question, a human answer, and a detailed multi-layer CoE rationale with explicit tags [2509.14851]. The paper states that SFT instills a predictable, interpretable reasoning scaffold, but can make outputs rigid without further adaptation [2509.14851].

During **Stage 2: Reinforcement Learning**, the framework applies **GRPO** to improve quality and empathy beyond structural adherence [2509.14851]. For each input, the model generates a group of candidate outputs; each output receives a composite reward; and the advantage for sample $i$ in the group is normalized as

$$
A_i = \frac{r_i - \mu_{\{r\}}}{\sigma_{\{r\}}}
$$

[2509.14851].

The reward model has two components. The **format reward** enforces compliance with the full CoE structure and is binary:

$$
r_\text{format} =
\begin{cases}
1 & \text{if format correct} \\
0 & \text{otherwise}
\end{cases}
$$

[2509.14851].

The **answer quality reward** is trained using contrastive learning. For a query-response pair, embeddings are compared through cosine similarity, and triplet loss separates professional responses from irrelevant, generic, or mismatched-emotion negatives:

$$
\mathcal{L}_{\text{contrastive}} = \max\left(0, \cos(E(q), E(a^-)) - \cos(E(q), E(a^+)) + m\right)
$$

[2509.14851].

During RL, if cosine similarity exceeds a threshold, $r_\text{answer}=1$; otherwise it is 0, and the total reward is

$$
r_\text{total} = r_\text{format} + r_\text{answer}
$$

[2509.14851].

This reward design differs from other empathetic RL systems. EmpRL defines empathy reward through agreement between generated and target empathy levels across emotional reaction, interpretation, and exploration, while also adding a KL penalty for coherence [2408.02976]. Kardia-R1 uses a unified reward combining format, emotion accuracy, and rubric reward with equal weights under GRPO [2512.01282]. Empathy-R1’s reward is narrower: structural correctness plus embedding-based answer quality [2509.14851]. This suggests a more constrained but task-focused alignment objective for long-form counseling responses.

## 5. Evaluation and empirical performance

Empathy-R1 is evaluated using both automatic metrics and human preference judgments [2509.14851]. The automatic metrics reported are **BLEU-1 (B-1), Distinct-1 (D-1), ROUGE-L (R-L), METEOR, and Normalized Average (NAvg)** [2509.14851]. The paper states that Empathy-R1 achieves **state-of-the-art in BLEU-1 and METEOR**, indicating strong semantic and lexical alignment [2509.14851].

The most prominent result is the human evaluation outcome. On the Empathy-QA test set, Empathy-R1 achieves a **Win@1 rate of 44.30%** [2509.14851]. On PsyQA, the reported Win@1 is **37.50%** [2509.14851]. The corresponding **Mean Rank** values are **2.60** on Empathy-QA and **2.62** on PsyQA [2509.14851]. The paper describes these results as showing clear preference over strong baselines including **Qwen3-8B without default thinking**, **CBT-LLM**, **EmoLLM**, and **Deepseek-R1** [2509.14851].

The paper also reports an ablation result on training strategy: **SFT alone yields rigid or less nuanced responses**, **RL alone is less effective without the SFT-initialized structure**, and the **combination yields best results** [2509.14851]. A further qualitative ablation states that removing CoE structure leads to decreased **BLEU-1** and **Distinct-1**, with responses becoming more repetitive and less insightful [2509.14851].

These findings align with a broader empirical pattern in empathetic dialogue research: structure plus explicit optimization tends to outperform either component in isolation. EmpRL reports that removing empathy reward or RL causes a steep drop in Emp-F1 and diversity [2408.02976]. Kardia-R1 reports stronger outcomes when rubric-based RL is layered on top of SFT [2512.01282]. ReEmpathy reports that gains disappear when reflective alternation is ablated [2601.18281]. Empathy-R1 extends this pattern to long-form mental health support [2509.14851].

## 6. Position within empathetic AI research

Empathy-R1 belongs to a family of systems that move beyond maximum-likelihood response generation toward explicit empathy modeling, structured reasoning, and alignment through task-specific reward signals. Earlier work such as IMAGINE integrated **emotion cause extraction**, **knowledge graph–based knowledge extension**, and **communication mechanism** prediction for empathetic response generation [2212.02000]. EmpRL made empathy level alignment itself the RL objective [2408.02976]. Kardia-R1 advanced interpretable, stepwise empathetic cognition through a rubric-LM judge and user-grounded data [2512.01282]. ReEmpathy introduced free-form reflective reasoning interleaved with spoken response generation [2601.18281]. Psyche-R1 unified empathy, expertise, and reasoning in a Chinese psychological LLM using SFT and GRPO [2508.10848].

Relative to these systems, Empathy-R1 is distinguished by four properties stated in the paper: its focus on **long counseling texts**, its **Chinese** setting, its **Chain-of-Empathy** scaffold, and its combination of **SFT plus GRPO** guided by a dedicated reward model [2509.14851]. Its interpretability claim is narrower than that of Kardia-R1, which explicitly segments outputs into understanding, reasoning, explicit emotion prediction, and empathetic response [2512.01282], but broader than systems where empathy remains latent and only the final response is observed.

The framework also reflects a methodological convergence with recent empathy evaluation research. Multi-dimensional evaluation work has argued that empathy should be assessed from both expressed and perceived perspectives rather than from expression alone [2402.11409]. ReEmpathy’s EmpathyEval similarly replaces single scalar signals with descriptive natural-language empathy assessment for spoken dialogue [2601.18281]. Empathy-R1 does not introduce a similarly multi-dimensional evaluator, but its use of an explicit rationale plus human preference assessment indicates the same pressure toward richer supervision and evaluation [2509.14851].

A common misconception in this area is that empathetic response generation is primarily a matter of producing emotionally warm phrasing. The literature summarized here suggests otherwise. IMAGINE emphasizes emotion cause and communication mechanism [2212.02000]. EmpRL distinguishes affective and cognitive empathy via emotional reaction, interpretation, and exploration [2408.02976]. Kardia-R1 adds user understanding, contextual reasoning, emotion perception, persona consistency, and safety [2512.01282]. Empathy-R1 contributes to this line by making emotions, causes, intentions, and response strategy explicit within long-form mental health support [2509.14851].

## 7. Significance, limitations, and implications

The significance claimed for Empathy-R1 is twofold: improved performance and increased trustworthiness through interpretability [2509.14851]. The framework is presented as enabling responses that are not boilerplate but instead reference specific user contexts and needs through four-layer analysis [2509.14851]. In this sense, CoE is not merely a prompt template; it is the structural premise on which training, reward design, and evaluation are organized.

The paper also frames Empathy-R1 as filling a gap in **Chinese-language mental health AI**, describing Empathy-QA as the largest and most relevant corpus for long counseling texts to date [2509.14851]. A plausible implication is that linguistic and cultural grounding are treated as part of empathetic competence rather than as peripheral deployment details.

At the same time, the reported design reveals several boundaries. The reward is based on **format correctness** and **embedding similarity thresholding** rather than on a richer rubric over empathy, safety, and persona consistency of the sort used in Kardia-R1 [2512.01282]. The evaluation emphasizes **Win@1**, **Mean Rank**, and lexical overlap metrics rather than multi-aspect human scoring. This suggests that Empathy-R1 prioritizes response preference and structural faithfulness, while leaving some dimensions of clinical or interpersonal evaluation less explicitly formalized.

More broadly, the framework illustrates a transition in empathetic AI from response imitation to reasoning-guided generation. This suggests a research trajectory in which empathy systems are increasingly expected to expose the intermediate stages by which they interpret emotion, attribute causes, infer user intent, and choose supportive strategies. Empathy-R1 exemplifies that trajectory for long-form mental health support in Chinese, using CoE as the organizing principle and reinforcement learning as the alignment mechanism [2509.14851].

Source: https://www.emergentmind.com/topics/empathy-r1