Discourse Diversity in Multi-Turn Empathic Dialogue

• URL: http://arxiv.org/abs/2604.11742v1
• Date: Mon, 13 Apr 2026 17:17:22 GMT
• Title: Discourse Diversity in Multi-Turn Empathic Dialogue
• Abstract summary: Large language models (LLMs) produce responses rated as highly empathic in single-turn settings.<n>LLMs are known to be formulaic generators that reuse the same lexical patterns, syntactic templates, and discourse structures across tasks.<n>We introduce MINT, the first reinforcement learning framework to optimize discourse move diversity across multi-turn empathic dialogue.
• Score: 27.43686763116342
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Large language models (LLMs) produce responses rated as highly empathic in single-turn settings (Ayers et al., 2023; Lee et al., 2024), yet they are also known to be formulaic generators that reuse the same lexical patterns, syntactic templates, and discourse structures across tasks (Jiang et al., 2025; Shaib et al., 2024; Namuduri et al., 2025). Less attention has been paid to whether this formulaicity extends to the level of discourse moves, i.e., what a response does for the person it is addressing. This question is especially consequential for empathic dialogue, where effective support demands not just a kind response at one moment but varied strategies as a conversation unfolds (Stiles et al., 1998). Indeed, prior work shows that LLMs reuse the same tactic sequences more than human supporters in single-turn settings (Gueorguieva et al., 2026). We extend this analysis to multi-turn conversations and find that the rigidity compounds: once a tactic appears in a supporter turn, LLMs reuse it in the next at nearly double the rate of humans (0.50-0.56 vs. 0.27). This pattern holds across LLMs serving as supporters in real emotional support conversations, and is invisible to standard similarity metrics. To address this gap, we introduce MINT (Multi-turn Inter-tactic Novelty Training), the first reinforcement learning framework to optimize discourse move diversity across multi-turn empathic dialogue. The best MINT variant combines an empathy quality reward with a cross-turn tactic novelty signal, improving aggregate empathy by 25.3% over vanilla across 1.7B and 4B models while reducing cross-turn discourse move repetition by 26.3% on the 4B model, surpassing all baselines including quality-only and token-level diversity methods on both measures. These results suggest that what current models lack is not empathy itself, but the ability to vary their discourse moves across a conversation.

Related papers

• AI generates well-liked but templatic empathic responses [26.705469030874223]
  People rate Large Language Models responses as more empathic than human-written responses.<n>We develop a taxonomy of 10 empathic language "tactics"<n>We find that LLM responses are highly formulaic at a discourse functional level.
  arXiv  Detail & Related papers  (2026-04-09T17:22:42Z)
• SCOPE: A Dataset of Stereotyped Prompts for Counterfactual Fairness Assessment of LLMs [45.94124349318317]
  SCOPE contains 241,280 prompts organized into 120,640 counterfactual pairs.<n>This resource provides a controlled, semantically aligned, and intent-aware basis for evaluating fairness, robustness, and counterfactual consistency.
  arXiv  Detail & Related papers  (2026-04-07T07:56:26Z)
• Reflecting Twice before Speaking with Empathy: Self-Reflective Alternating Inference for Empathy-Aware End-to-End Spoken Dialogue [53.95386201009769]
  We introduce EmpathyEval, a descriptive natural-language-based evaluation model for assessing empathetic quality in spoken dialogues.<n>We propose ReEmpathy, an end-to-end Spoken Language Models that enhances empathetic dialogue through a novel Empathetic Self-Reflective Alternating Inference mechanism.
  arXiv  Detail & Related papers  (2026-01-26T09:04:50Z)
• PanoSent: A Panoptic Sextuple Extraction Benchmark for Multimodal Conversational Aspect-based Sentiment Analysis [74.41260927676747]
  This paper bridges the gaps by introducing a multimodal conversational Sentiment Analysis (ABSA) To benchmark the tasks, we construct PanoSent, a dataset annotated both manually and automatically, featuring high quality, large scale, multimodality, multilingualism, multi-scenarios, and covering both implicit and explicit sentiment elements. To effectively address the tasks, we devise a novel Chain-of-Sentiment reasoning framework, together with a novel multimodal large language model (namely Sentica) and a paraphrase-based verification mechanism.
  arXiv  Detail & Related papers  (2024-08-18T13:51:01Z)
• The MuSe 2024 Multimodal Sentiment Analysis Challenge: Social Perception and Humor Recognition [64.5207572897806]
  The Multimodal Sentiment Analysis Challenge (MuSe) 2024 addresses two contemporary multimodal affect and sentiment analysis problems. In the Social Perception Sub-Challenge (MuSe-Perception), participants will predict 16 different social attributes of individuals. The Cross-Cultural Humor Detection Sub-Challenge (MuSe-Humor) dataset expands upon the Passau Spontaneous Football Coach Humor dataset.
  arXiv  Detail & Related papers  (2024-06-11T22:26:20Z)
• LastResort at SemEval-2024 Task 3: Exploring Multimodal Emotion Cause Pair Extraction as Sequence Labelling Task [3.489826905722736]
  SemEval 2024 introduces the task of Multimodal Emotion Cause Analysis in Conversations. This paper proposes models that tackle this task as an utterance labeling and a sequence labeling problem. In the official leaderboard for the task, our architecture was ranked 8th, achieving an F1-score of 0.1759 on the leaderboard.
  arXiv  Detail & Related papers  (2024-04-02T16:32:49Z)
• MT-Bench-101: A Fine-Grained Benchmark for Evaluating Large Language Models in Multi-Turn Dialogues [58.33076950775072]
  MT-Bench-101 is designed to evaluate the fine-grained abilities of Large Language Models (LLMs) in multi-turn dialogues. We construct a three-tier hierarchical ability taxonomy comprising 4208 turns across 1388 multi-turn dialogues in 13 distinct tasks. We then evaluate 21 popular LLMs based on MT-Bench-101, conducting comprehensive analyses from both ability and task perspectives.
  arXiv  Detail & Related papers  (2024-02-22T18:21:59Z)
• SoulChat: Improving LLMs' Empathy, Listening, and Comfort Abilities through Fine-tuning with Multi-turn Empathy Conversations [19.11368665202549]
  When large language models are applied in the field of psychological counseling, they often rush to provide universal advice. We constructed a multi-turn empathetic conversation dataset of more than 2 million samples. Experiments have shown that the empathy ability of LLMs can be significantly enhanced when finetuning by using multi-turn dialogue history.
  arXiv  Detail & Related papers  (2023-11-01T03:49:52Z)
• BotChat: Evaluating LLMs' Capabilities of Having Multi-Turn Dialogues [72.65163468440434]
  This report provides a preliminary evaluation of existing large language models for human-style multi-turn chatting. We prompt large language models (LLMs) to generate a full multi-turn dialogue based on the ChatSEED, utterance by utterance. We find GPT-4 can generate human-style multi-turn dialogues with impressive quality, significantly outperforms its counterparts.
  arXiv  Detail & Related papers  (2023-10-20T16:53:51Z)
• Generating Dialogue Responses from a Semantic Latent Space [75.18449428414736]
  We propose an alternative to the end-to-end classification on vocabulary. We learn the pair relationship between the prompts and responses as a regression task on a latent space. Human evaluation showed that learning the task on a continuous space can generate responses that are both relevant and informative.
  arXiv  Detail & Related papers  (2020-10-04T19:06:16Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.