Why Qwen is Dynamically Stable: An Empirical Phase Map of 10 LLMs
I’ve been analyzing the internal trajectory dynamics of several open-source models (Qwen, Llama, Gemma, etc.) using a new metric called ct_t (local trajectory instability based on hidden state curvature).
The results show a clear structural difference in how these models process information internally. While models like Gemma-2B often drift into a "Chaotic" regime (high instability spikes) and Llama-3.2 into an "Underactive" one (rigid, low variance), Qwen models consistently maintain an "Adaptive" regime.
This suggests that Qwen’s architecture achieves a unique balance between stability and flexibility, independent of model size. A 1.5B Qwen model was dynamically more stable than larger counterparts in our panel. The 4 Regimes Identified:
Underactive: Rigid, low adaptability.
Adaptive: Balanced flux/stability (Qwen's zone).
Transition: Boundary zone.
Chaotic: High instability, prone to divergence. I’ve published the full working paper with data from 158 runs on Zenodo:
Four Dynamical Regimes in Large Language Models: An Empirical Phase Map
I’d love to hear your thoughts: Does this "Adaptive" dynamic correlate with your experience of Qwen's reasoning capabilities?
Best,
Jean-Denis Bosange Batuli