🧠 Evolved Qwen Hybrid

Technical Overview

• Core LLM: Qwen 3 (0.6B) - Optimized for high-speed reasoning.
• Memory Engine: SELSC v2 (Synaptic Evolved Learning State Controller).
• Learning Logic: Online Hebbian learning with custom dopamine-gated reward heads.
• Architecture: SNN-LLM Hybrid with dynamic synaptic growth (Neurogenesis).

This project explores the intersection of transformer architectures and spiking neural networks, aiming to solve catastrophic forgetting in localized training environments.