LLM Modules: Knowledge Transfer from a Large to a Small Model using Enhanced Cross-Attention
Research paper by Konstantin Kolomeitsev. Published 2025-02-12. Topics: LLM Modules, Enhanced Cross-Attention, Qwen2-1.5B, GPT-Neo-125M, Bespoke-Stratos-17k dataset.
Resource information
- Authors
- Konstantin Kolomeitsev
- Research area
- Language technology
- Application domain
- General AI research
- AI methods
- Large language models
- Research topics
- cs.CL, cs.LG
- arXiv record date
- 2025-02-12
- Publication date
- 2025-02-12
- Publication
- arXiv
- Upstream last modified
- Feb 13, 2025
- License
- CC BY 4.0
- Last record update
- Nov 3, 2026
Abstract
In this work, we propose an architecture of LLM Modules that enables the transfer of knowledge from a large pre-trained model to a smaller model using an Enhanced Cross-Attention mechanism. In the proposed scheme, the Qwen2-1.5B model is frozen and its representations are passed through specially designed attention layers to the GPT-Neo-125M model, which is trained on limited computational resources. Experimental results on the Bespoke-Stratos-17k dataset demonstrate that after 15 epochs of training, the combined model generates responses comparable in quality to those obtained by distillation. We discuss the advantages of the modular approach, provide examples of input queries and comparative analysis, and outline prospects for further extension of the method.