base_model:
- google/gemma-3-4b-it
license: mit
library_name: transformers
pipeline_tag: text-generation
Sequential-Scaled-Planner-Gemma3-4B
🌐 Project Page | 💻 Code | 📄 Paper
We introduce RecursiveMAS, a multi-agent framework 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 Sequential-Scaled setting, the Planner Agent is responsible for producing the initial reasoning plan, which is then passed to subsequent agents through lightweight RecursiveLink modules for further refinement and solution generation.
Model Details
| Item | Description |
|---|---|
| Model | Sequential-Scaled-Planner-Gemma3-4B |
| Collaboration Style | Sequential-Scaled |
| Agent Role | Planner Agent |
| Base Model | Gemma3-4B |
⚠️ Note: This checkpoint is a role-specific agent in RecursiveMAS, rather than a standalone model intended for plain-text generation.
For detailed usage instructions, please refer to our GitHub repository.
Sample Usage
To use this model within the RecursiveMAS framework, you can load the entire system as follows:
from system_loader import load_mas_system
# Load the whole MAS pipeline
mas = load_mas_system(
style="sequential_scaled",
device="cuda",
trust_remote_code=True,
)
planner = mas.agents["planner"].model
critic = mas.agents["critic"].model
solver = mas.agents["solver"].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},
}