Papers
arxiv:2510.05781

Mixture of Neuron Experts

Published on Oct 7, 2025
Authors:
,
,
,
,
,
,
,
,

Abstract

Mixture of Neuron Experts enables sparse neuron-level selection within MoE layers to halve activated parameters without sacrificing performance or adding routing overhead.

In this work, we first explore whether the parameters activated by the MoE layer remain highly sparse at inference. We perform a sparsification study on several representative MoE models. For each expert, we rank parameters by the magnitude of their activations from the gate projection and progressively prune the activated subset. Pruning up to 60% of parameters within that subset causes only negligible task-performance degradation; substantial drops occur only after more than 90% are removed. We further decompose experts into neuron-granular MoE and visualize their activation values, finding that most neuron activations are near zero. This observation motivates us to select only high-activation neuron experts during pretraining. Based on this insight, we propose Mixture of Neuron Experts (MoNE). MoNE achieves neuron-granular expert selection by only applying a simple top-k selection within each expert, incurs negligible latency, and requires no additional routing parameters or inter-expert communication. Extensive experiments demonstrate that MoNE matches traditional MoE performance while activating only 50% of the MoE-layer parameters, and it consistently outperforms traditional MoE when compared at equal numbers of activated parameters. These results suggest that MoNE is a practical approach to improving parameter utilization and inference efficiency in MoE-like models.

Community

Sign up or log in to comment

Get this paper in your agent:

hf papers read 2510.05781
Don't have the latest CLI?
curl -LsSf https://hf.co/cli/install.sh | bash

Models citing this paper 0

No model linking this paper

Cite arxiv.org/abs/2510.05781 in a model README.md to link it from this page.

Datasets citing this paper 0

No dataset linking this paper

Cite arxiv.org/abs/2510.05781 in a dataset README.md to link it from this page.

Spaces citing this paper 0

No Space linking this paper

Cite arxiv.org/abs/2510.05781 in a Space README.md to link it from this page.

Collections including this paper 0

No Collection including this paper

Add this paper to a collection to link it from this page.