A Systematic Study of In-the-Wild Model Merging for Large Language Models

TL;DR

This paper systematically evaluates various model merging techniques for large language models trained on overlapping or conflicting objectives, revealing that simple task arithmetic often outperforms more complex methods in real-world scenarios.

Contribution

It provides a large-scale, empirical comparison of six merging methods across multiple LLMs and benchmarks, highlighting the effectiveness of simple task arithmetic in heterogeneous settings.

Findings

Task Arithmetic reliably improves performance in in-the-wild merging scenarios.

Most advanced merging methods do not outperform the base models in heterogeneous settings.

Current merging techniques struggle to extract useful updates from conflicting models.

Abstract

Model merging combines multiple fine-tuned checkpoints into a single model without additional training, offering an attractive approach to reusing models and efficiently improving performance. However, it remains unclear whether the advantages reported for settings where all merged experts have distinct roles and are tuned on clearly separated tasks also hold in settings where the merged experts do not have clearly distinct roles, but are trained on overlapping or even conflicting objectives. To evaluate this setting, we present a large-scale, systematic evaluation of "in-the-wild" model merging of heterogeneous experts, that may have been trained on overlapping or conflicting objectives. Concretely, we evaluate six state-of-the-art merging methods, including recent subspace methods, across four open-weight LLMs, twelve fine-tuned checkpoints per base model, and sixteen standard LLM benchmarks. Evaluating through standardized benchmarks, we measure both the probability that a model merged from a heterogeneous set of experts outperforms the base model and we measure relative gains over the best individual checkpoint. Our results show that the oldest and simplest method, Task Arithmetic, is the only approach that reliably yields performance gains on LLMs in this "in-the-wild" setting. Other interference-aware and subspace merging methods typically do not result in notable improvements over the base model. Our findings indicate that current merging techniques mostly do not enable extracting useful weight updates from heterogeneous and potentially conflicting versions. This motivates the design of LLM-specific merging algorithms and merging-aware fine-tuning methods.