Mind2Dialogue: Training Human-Aware Language Models by Simulating User Mental States

• URL: http://arxiv.org/abs/2609.15972v1
• Date: Mon, 14 Sep 2026 17:55:58 GMT
• Title: Mind2Dialogue: Training Human-Aware Language Models by Simulating User Mental States
• Abstract summary: Mind2Dialogue is a framework for simulating users' mental states and turning them into privileged supervision for human-aware training.<n>We first propose a psychology-guided simulator that preserves personal characteristics while updating mental states to generate coherent conversations.<n>Our privileged distillation then trains models on the Oracle's well-informed responses to assist users without direct access to their mental states at deployment.
• Score: 72.77158276658636
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: As language models become more capable, long-term collaboration in learning, reasoning, and decision-making calls for a deeper understanding of the people they serve. Yet training such human-aware language models faces a fundamental supervision gap because current datasets for LLM assistant training contain few if any well-informed responses explicitly grounded in users' unspoken beliefs and goals. Scaling such supervision is inherently constrained, as users' underlying states are not directly observable. We thus propose the Mind2Dialogue framework to mitigate this gap by simulating users' mental states and turning them into privileged supervision for human-aware training. Specifically, we first propose a psychology-guided simulator that preserves personal characteristics while updating mental states through interaction to generate coherent conversations. The key idea is to enforce a shared evolving mental state that drives user behavior and guides an Oracle assistant's responses. Our privileged distillation then trains models on the Oracle's well-informed responses to assist users without direct access to their mental states at deployment. Moreover, we propose to evaluate human-aware learning by combining personalization and theory of mind, examining how models understand people and act on that understanding. Training on the full Mind2Dialogue corpus improves every reported personalization metric over the corresponding Qwen, Llama, and OLMo instruction-tuned baselines, including gains of 26.6 to 40.9 percentage points in preference-following generation. The gains extend to belief and action reasoning on Qwen and Llama, beyond personalized assistance. Looking forward, Mind2Dialogue makes user simulation a foundation for genuine AI collaborators that understand beliefs and intentions behind people's words and support their long-term goals across education, work, and everyday life.

Related papers

• Measuring User's Mental Models of Speech Translation in Human-AI Collaboration [23.64422798744437]
  This paper studies users' mental models of speech translation systems through a new framework based on cross-lingual question answering.<n>By analyzing user behavior and accuracy trends across varying translation qualities, we examine to what extent they can predict where the system is likely to be wrong.
  arXiv  Detail & Related papers  (2026-06-23T14:40:36Z)
• Resonant Minds: Closed-Loop Social Avatars with Theory of Mind [16.880605576970538]
  We propose a closed-loop dual-agent framework integrating perception, social reasoning, and expression into a continuous interaction cycle.<n>The perception module analyzes partners' multimodal behaviors from video, while the social reasoning module infers hidden mental states through Theory of Mind.<n>The expression module then generates emotion-controllable dual-agent videos both speaker speech and expression alongside listener reactive behaviors.
  arXiv  Detail & Related papers  (2026-06-04T09:03:43Z)
• ThoughtTrace: Understanding User Thoughts in Real-World LLM Interactions [31.70995208851302]
  ThoughtTrace is the first large-scale dataset that pairs real-world multi-turn human--AI conversations with users' self-reported thoughts.<n>ThoughtTrace comprises 1,058 users, 2,155 conversations, 17,058 turns, and 10,174 thought annotations collected across 20 language models.
  arXiv  Detail & Related papers  (2026-05-19T16:42:06Z)
• HumanLLM: Towards Personalized Understanding and Simulation of Human Nature [72.55730315685837]
  HumanLLM is a foundation model designed for personalized understanding and simulation of individuals.<n>We first construct the Cognitive Genome, a large-scale corpus curated from real-world user data on platforms like Reddit, Twitter, Blogger, and Amazon.<n>We then formulate diverse learning tasks and perform supervised fine-tuning to empower the model to predict a wide range of individualized human behaviors, thoughts, and experiences.
  arXiv  Detail & Related papers  (2026-01-22T09:27:27Z)
• Because we have LLMs, we Can and Should Pursue Agentic Interpretability [22.10895793309226]
  Large Language Models (LLMs) proactively assist human understanding by developing and leveraging a mental model of the user.<n>Agentic interpretability introduces challenges, particularly in evaluation, due to what we call human-entangled-in-the-loop' nature.<n>Agentic interpretability's promise is to help humans learn the potentially deceptive concepts of the LLMs, rather than see us fall increasingly far from understanding them.
  arXiv  Detail & Related papers  (2025-06-13T18:13:58Z)
• Training Language Models for Social Deduction with Multi-Agent Reinforcement Learning [31.196865401472664]
  We train language models to have productive discussions about their environment in natural language without any human demonstrations.<n>We leverage the agent's goal to predict useful information about the world as a dense reward signal that guides communication.<n>We analyze emergent behaviors due to our technique, such as accusing suspects and providing evidence, and find that it enables strong discussions.
  arXiv  Detail & Related papers  (2025-02-09T22:44:45Z)
• Interactive Dialogue Agents via Reinforcement Learning on Hindsight Regenerations [58.65755268815283]
  Many real dialogues are interactive, meaning an agent's utterances will influence their conversational partner, elicit information, or change their opinion. We use this fact to rewrite and augment existing suboptimal data, and train via offline reinforcement learning (RL) an agent that outperforms both prompting and learning from unaltered human demonstrations. Our results in a user study with real humans show that our approach greatly outperforms existing state-of-the-art dialogue agents.
  arXiv  Detail & Related papers  (2024-11-07T21:37:51Z)
• Real-time Addressee Estimation: Deployment of a Deep-Learning Model on the iCub Robot [52.277579221741746]
  Addressee Estimation is a skill essential for social robots to interact smoothly with humans. Inspired by human perceptual skills, a deep-learning model for Addressee Estimation is designed, trained, and deployed on an iCub robot. The study presents the procedure of such implementation and the performance of the model deployed in real-time human-robot interaction.
  arXiv  Detail & Related papers  (2023-11-09T13:01:21Z)
• AGENT: A Benchmark for Core Psychological Reasoning [60.35621718321559]
  Intuitive psychology is the ability to reason about hidden mental variables that drive observable actions. Despite recent interest in machine agents that reason about other agents, it is not clear if such agents learn or hold the core psychology principles that drive human reasoning. We present a benchmark consisting of procedurally generated 3D animations, AGENT, structured around four scenarios.
  arXiv  Detail & Related papers  (2021-02-24T14:58:23Z)
• A Neural Topical Expansion Framework for Unstructured Persona-oriented Dialogue Generation [52.743311026230714]
  Persona Exploration and Exploitation (PEE) is able to extend the predefined user persona description with semantically correlated content. PEE consists of two main modules: persona exploration and persona exploitation. Our approach outperforms state-of-the-art baselines in terms of both automatic and human evaluations.
  arXiv  Detail & Related papers  (2020-02-06T08:24:33Z)

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.