1. Tutorial Introduction

UserLM-8b is a user behavior simulation model released by Microsoft in October 2025. Unlike typical LLMs that play the role of "assistant" in conversations, UserLM-8b is designed to simulate the "user" role in conversations (trained on the WildChat dialogue corpus) and can be used for evaluating the capabilities of large-scale assistant models. This model is not a common large-scale assistant model and cannot simulate more realistic conversations or problem-solving, but it helps in developing more powerful assistants. Related research papers are available. Flipping the Dialogue: Training and Evaluating User Language Models .

The computing resources used in this tutorial are dual RTX 5090 cards.

2. Effect Examples

3. Operation steps

1. Start the container

2. After entering the webpage, you can start a conversation with the model

If "Bad Gateway" is displayed, it means the model is initializing. Since the model is large, please wait about 2-3 minutes and refresh the page.

How to use 

Citation Information

@inproceedings{naous2025flipping,
  title={Flipping the Dialogue: Training and Evaluating User Language Models},
  author={Tarek Naous and Philippe Laban and Wei Xu and Jennifer Neville},
  journal={arXiv preprint arXiv:2510.06552},
  year={2025},
  url={https://arxiv.org/abs/2510.06552}
}