[AAAI 2026] SmartAgent: Chain-of-User-Thought for Embodied Personalized Agent in Cyber World

The official repository of our paper "SmartAgent: Chain-of-User-Thought for Embodied Personalized Agent in Cyber World".

Blogs 🤗

• Find SmartAgent in Awesome-LLM-Reasoning with more brilliant works on LLM/MLLM reasoning!
• Find SmartAgent in Multimodal Chain-of-Thought Reasoning: A Comprehensive Survey with the latest Multimodal CoT reasoning techniques toward multimodal AGI!
• Find SmartAgent in Agentic Web: Weaving the Next Web with AI Agents opening opportunities and challenges of intelligence, interaction, and economics for the next-generation web technologies!
• Find SmartAgent in LLM-Agent-for-Recommendation-and-Search with more fascinating LLM agents inspiring the next-generation information retrieval!

Chain-of-User-Thought (COUT) Reasoning Paradigm

We formulate COUT to achieve embodied personalized agent training in terms of three stages of thought. In Thought #1, according to a user's instruction, an agent performs GUI actions to search for an item pool. In Thought #2 with seeing the pool, the agent reasons underlying requirements behind the original instruction, as implied by the previous actions. In Thought #3, based on the underlying thought, the agent recommends items within the pool to complete the user's instruction.

SmartSpot Benchmark

Download Link

• Google Drive: image data, annotation data (zh), annotation data (en).

Datasets Statistics

Requirements

Install the Python dependencies:

pip install -r requirements.txt

Run SmartAgent

Data Preprocessing

python process.py

Training

bash run.sh

Evaluation

python test.py

Convert LoRA Checkpoints

python resave_ckpt.py

Citation

Please consider citing our paper and staring this repo if you find SmartAgent helpful in your work, thanks!

@inproceedings{zhang2026smartagent,
  title={Smartagent: Chain-of-user-thought for embodied personalized agent in cyber world},
  author={Zhang, Jiaqi and Gao, Chen and Zhang, Liyuan and Nguyen, Quoc Viet Hung and Yin, Hongzhi},
  booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},
  volume={40},
  number={21},
  pages={17993--18001},
  year={2026}
}