TL;DR
This paper investigates the lack of discourse move diversity in large language models during multi-turn empathic dialogues and introduces a reinforcement learning framework, MINT, to enhance move variety and empathy.
Contribution
It reveals the extent of discourse move repetition in LLMs and proposes MINT, a novel reinforcement learning method to improve move diversity and empathy in multi-turn conversations.
Findings
LLMs reuse discourse tactics nearly twice as often as humans.
MINT reduces cross-turn tactic repetition by 26.3%.
MINT improves empathy scores by 25.3% over baseline models.
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…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Code & Models
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
