Research

Paper

AI LLM March 13, 2026

Expert Pyramid Tuning: Efficient Parameter Fine-Tuning for Expertise-Driven Task Allocation

Authors

Jia-Chen Zhang, Zhen-Wei Yan, Yu-Jie Xiong, Chun-Ming Xia

Abstract

Parameter-Efficient Fine-Tuning (PEFT) has become a dominant paradigm for deploying LLMs in multi-task scenarios due to its extreme parameter efficiency. While Mixture-of-Experts (MoE) based LoRA variants have achieved promising results by dynamically routing tokens to different low-rank experts, they largely overlook the hierarchical nature of task complexity. Existing methods typically employ experts with uniform architectures, limiting their ability to capture diverse feature granularities required by distinct tasks--where some tasks demand high-level semantic abstraction while others require fine-grained syntactic manipulation. To bridge this gap, we propose Expert Pyramid Tuning (EPT), a novel architecture that integrates the multi-scale feature pyramid concept from computer vision into the realm of PEFT. Unlike standard LoRA, EPT decomposes task adaptation into two stages: (1) A shared meta-knowledge Subspace that encodes universal linguistic patterns in low dimensions; (2) A Pyramid Projection Mechanism that utilizes learnable up-projection operators to reconstruct high-dimensional features at varying scales. A task-aware router then dynamically selects the optimal combination of these multi-scale features. Extensive experiments across multiple multi-task benchmarks demonstrate that EPT significantly outperforms SOTA MoE-LoRA variants. Crucially, thanks to the re-parameterization capability of our design, EPT achieves this performance improvement while simultaneously reducing the number of training parameters.

Metadata

arXiv ID: 2603.12577
Provider: ARXIV
Primary Category: cs.CL
Published: 2026-03-13
Fetched: 2026-03-16 06:01

Related papers

Raw Data (Debug)
{
  "raw_xml": "<entry>\n    <id>http://arxiv.org/abs/2603.12577v1</id>\n    <title>Expert Pyramid Tuning: Efficient Parameter Fine-Tuning for Expertise-Driven Task Allocation</title>\n    <updated>2026-03-13T02:23:41Z</updated>\n    <link href='https://arxiv.org/abs/2603.12577v1' rel='alternate' type='text/html'/>\n    <link href='https://arxiv.org/pdf/2603.12577v1' rel='related' title='pdf' type='application/pdf'/>\n    <summary>Parameter-Efficient Fine-Tuning (PEFT) has become a dominant paradigm for deploying LLMs in multi-task scenarios due to its extreme parameter efficiency. While Mixture-of-Experts (MoE) based LoRA variants have achieved promising results by dynamically routing tokens to different low-rank experts, they largely overlook the hierarchical nature of task complexity. Existing methods typically employ experts with uniform architectures, limiting their ability to capture diverse feature granularities required by distinct tasks--where some tasks demand high-level semantic abstraction while others require fine-grained syntactic manipulation. To bridge this gap, we propose Expert Pyramid Tuning (EPT), a novel architecture that integrates the multi-scale feature pyramid concept from computer vision into the realm of PEFT. Unlike standard LoRA, EPT decomposes task adaptation into two stages: (1) A shared meta-knowledge Subspace that encodes universal linguistic patterns in low dimensions; (2) A Pyramid Projection Mechanism that utilizes learnable up-projection operators to reconstruct high-dimensional features at varying scales. A task-aware router then dynamically selects the optimal combination of these multi-scale features. Extensive experiments across multiple multi-task benchmarks demonstrate that EPT significantly outperforms SOTA MoE-LoRA variants. Crucially, thanks to the re-parameterization capability of our design, EPT achieves this performance improvement while simultaneously reducing the number of training parameters.</summary>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.CL'/>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.CV'/>\n    <published>2026-03-13T02:23:41Z</published>\n    <arxiv:primary_category term='cs.CL'/>\n    <author>\n      <name>Jia-Chen Zhang</name>\n    </author>\n    <author>\n      <name>Zhen-Wei Yan</name>\n    </author>\n    <author>\n      <name>Yu-Jie Xiong</name>\n    </author>\n    <author>\n      <name>Chun-Ming Xia</name>\n    </author>\n  </entry>"
}