0

MoD: A Distribution-Based Approach for Merging Large Language Models

A Mixture of Distributions (MoD) framework effectively merges large language models by combining their output probability distributions, preserving specialized capabilities and improving knowledge sharing across tasks.

Year
2024
Venue
arXiv 2024
Authors
2
Hosting
Abstract onlyARXIV-DEFAULT

Cite

Notes

Only stored in your browser.

Attribution

Abstract & full text
arxiv.org/abs/2411.00406ARXIV-DEFAULT
TL;DR
Semantic Scholar
Attribution policy →

Abstract

Large language models (LLMs) have enabled the development of numerous specialized, task-specific variants. However, the maintenance and deployment of these individual models present substantial challenges in terms of resource utilization and operational efficiency. In this work, we propose the \textit{Mixture of Distributions (MoD)} framework, a novel approach for merging LLMs that operates directly on their output probability distributions, rather than on model weights. Unlike traditional weight-averaging methods, MoD effectively preserves the specialized capabilities of individual models while enabling efficient knowledge sharing across tasks. Through extensive experimentation on mathematical reasoning benchmarks using Qwen2.5 models, we demonstrate that MoD significantly outperforms existing model merging techniques across multiple benchmarks. All code, data, and experimental materials are published at https://github.com/knovel-eng/mod.

Authors

2