Model merging, a method that combines the parameters and embeddings of multiple fine-tuned large language models (LLMs), offers a promising approach to enhance model performance across various tasks while maintaining computational efficiency. This paper introduces Activation-Informed Merging (AIM), a technique that integrates the information from the activation space of LLMs into the merging process to improve performance and robustness. AIM is designed as a flexible, complementary solution that is applicable to any existing merging method. It aims to preserve critical weights from the base model, drawing on principles from continual learning~(CL) and model compression. Utilizing a task-agnostic calibration set, AIM selectively prioritizes essential weights during merging. We empirically demonstrate that AIM significantly enhances the performance of merged models across multiple benchmarks. Our findings suggest that considering the activation-space information can provide substantial advancements in the model merging strategies for LLMs with up to 40\% increase in benchmark performance.
View on arXiv@article{nobari2025_2502.02421,
title={ Activation-Informed Merging of Large Language Models },
author={ Amin Heyrani Nobari and Kaveh Alimohammadi and Ali ArjomandBigdeli and Akash Srivastava and Faez Ahmed and Navid Azizan },
journal={arXiv preprint arXiv:2502.02421},
year={ 2025 }
}