base_model:
- Qwen/Qwen3.5-2B
license: mit
pipeline_tag: text-generation
library_name: transformers
Mixture-Summarizer-Qwen3.5-2B
🌐 Project Page | 💻 Code | 📄 Paper
We introduce RecursiveMAS, a multi-agent framework introduced in Recursive Multi-Agent Systems that scales agent collaboration through latent-space recursion.
RecursiveMAS treats a multi-agent system as a unified recursive computation, where heterogeneous agents iteratively exchange, refine, and evolve their latent states across recursion rounds. In the Mixture-Style setting, this Summarizer Agent integrates outputs from domain-specialized agents and produces the final response through recursive latent-space collaboration.
Model Details
| Item | Description |
|---|---|
| Model | Mixture-Summarizer-Qwen3.5-2B |
| Collaboration Style | Mixture-Style |
| Agent Role | Summarizer Agent |
| Base Model | Qwen3.5-2B |
⚠️ Note: This checkpoint is a role-specific agent in RecursiveMAS, rather than a standalone model intended for plain-text generation.
Usage
To use this agent as part of the RecursiveMAS pipeline, you can load the system using the provided loader from the official repository:
from system_loader import load_mas_system
# Load the whole Mixture-Style MAS pipeline
mas = load_mas_system(
style="mixture",
device="cuda",
trust_remote_code=True,
)
# Access the specific summarizer model
summarizer = mas.agents["summarizer"].model
Model Collections for RecursiveMAS
| Style | Model Collection |
|---|---|
| Sequential-Style | 🤗 HuggingFace |
| Mixture-Style | 🤗 HuggingFace |
| Distillation-Style | 🤗 HuggingFace |
| Deliberation-Style | 🤗 HuggingFace |
Experiment Results
Citation
@misc{recursivemas,
title={Recursive Multi-Agent Systems},
author={Xiyuan Yang and Jiaru Zou and Rui Pan and Ruizhong Qiu and Pan Lu and Shizhe Diao and Jindong Jiang and Hanghang Tong and Tong Zhang and Markus J. Buehler and Jingrui He and James Zou},
year={2026},
eprint={2604.25917},
archivePrefix={arXiv},
primaryClass={cs.AI},
url={https://arxiv.org/abs/2604.25917},
}