Real Time
Alibaba PAI Open-Sources AgenticQwen: "Dual Data Flywheels" Propel 8B Model to Rival 235B Performanc
Alibaba’s Platform for AI (PAI) team has officially released and open-sourced AgenticQwen, a new series of Small Language Models specifically eng...
3 weeks ago
•
April 27, 2026
•
22 views
Traditional synthetic data methods often suffer from homogeneity, causing Model Performance to plateau. AgenticQwen overcomes this limitation through a dynamic training APProach involving two distinct flywheels: Reasoning Flywheel: This mechanism automatically generates harder variants of problems based on the model's previous errors, forcing continuous improvement in logic and reasoning.
Agent Flywheel: Instead of simple Linear workflows (e.g., a straightforward booking process), this flywheel expands execution trajectories into complex behavior trees. It simulates real-world decision-making by introducing constraints, rejection scenarios, and adversarial conditions.
evaluations indicate that AgenticQwen delivers exceptional results on real-world tool-use benchmarks, such as TAU-2 and BFCL-V4: AgenticQwen-8B: Achieved an average score of 47.4, vastly outperforming the base Qwen3-8B (23.8) and closely approaching the performance of the massive Qwen3-235B (52.0).
AgenticQwen-30B-A3B: Utilizing a Mixture of Experts (MoE) architecture that ACTivates only 3B parameters during inference, this model reached a score of 50.2.
Industrial Application & Limitations
The model has already been deployed in internal production systems similar to manus, dEMOnstrating a significant reduction in end-to-end inference time compared to larger models. However, the team notes that due to a native context window of 40K tokens, small models like AgenticQwen still face limitations in deep search tasks requiring extensive context retention.
Be the first to rate this article.
Comments & Questions (0)
No comments yet
Be the first to comment!