首页 > AI前沿 > From Experts to Sub-experts: Fine-grained Parameter-Efficient Fine-Tuning for MoE LLMs

From Experts to Sub-experts: Fine-grained Parameter-Efficient Fine-Tuning for MoE LLMs

arXiv机器学习 2026-09-22 12:04 7 阅读 查看原文

As large language models (LLMs) scale rapidly, dense full-parameter adaptation becomes increasingly expensive, motivating sparse and modular architectures such as Mixture-of-Experts (MoE) models.

This shift raises a key question for parameter-efficient fine-tuning (PEFT): at what granularity should parameters be selected and updated?

Existing PEFT methods such as LoRA operate on predefined weight matrices, while expert-level sparse tuning methods update entire selected experts.

However, we observe that activated experts are internally sparse, with only a small fraction of intermediate channels strongly responding to downstream tasks, indicating that expert-level adaptation is still too coarse.

We propose NSFT (Neural Sub-expert Fine-Tuning), a fine-grained PEFT framework that refines MoE adaptation from experts to sub-experts.

NSFT decomposes each expert along the intermediate dimension into structured channel groups and selects task-relevant sub-experts by combining routing importance with intra-expert activation saliency.

To optimize sparse partial updates, NSFT further introduces learning-rate scaling and dynamic gradient scaling to compensate for the reduced effective update magnitude.

Experiments on OLMoE and Ling-mini-2.0 across challenging domain-specific tasks and general benchmarks show that NSFT consistently outperforms representative PEFT and expert-level sparse tuning baselines, while using substantially fewer trainable parameters and preserving competitive general capability.

These results suggest that sub-expert-level adaptation is a more precise and efficient PEFT paradigm for MoE LLMs.