AI Paper Atlas

Research map / Architectures and efficiency / Mixture of experts

Mosaic Pruning: A Hierarchical Framework for Generalizable Pruning of Mixture-of-Experts Models

Wentao Hu, Mingkuan Zhao, Shuangyong Song, Xiaoyan Zhu, Xin Lai, Jiayin Wang

AAAI 2026 · Architectures and efficiency › Mixture of experts · approach: Efficient MoE language model systems · 8 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add global activation-variability selection (GVP) over reconstruction-only Enumeration (ablation)Enumeration Pruning · 7 general benchmarks (Mixtral, 6 experts)average accuracy: 72.79 -> 73.50Improved
add cluster-then-select on top of GVP (full MoP) (ablation)GVP · 7 general benchmarks (Mixtral, 6 experts)average accuracy: 73.50 -> 74.01Improved
add cluster-then-select on top of GVP on Qwen (ablation)GVP · 7 general benchmarks (Qwen, 50 experts)average accuracy: 55.56 -> 59.10Improved
use mixed-diversity calibration data instead of C4 for Enumeration Pruning (ablation)Enumeration Pruning · 7 general benchmarks (Qwen, 50 experts)average accuracy: 55.11 -> 49.47Hurt
use mixed calibration data instead of C4 for Enumeration on Mixtral (ablation)Enumeration Pruning · 7 general benchmarks (Mixtral, 6 experts)average accuracy: 72.73 -> 72.79No clear effect
use MoP instead of Enumeration Pruning at 4 expertsexpert pruning · 7 general benchmarks (Mixtral, 4 experts)average accuracy: 62.41 -> 68.96Improved
use MoP instead of Enumeration Pruning at 40 expertsexpert pruning · 7 general benchmarks (Qwen, 40 experts)average accuracy: 38.60 -> 55.74Improved
use MoP instead of Enumeration Pruning on code at 4 expertsexpert pruning · HumanEval (Mixtral, 4 experts)pass rate: 1.22 -> 71.34Improved
use MoP on MATH, Mixtral 4 expertsexpert pruning · MATH (Mixtral, 4 experts)accuracy: 3.80 -> 3.00Hurt
use MoP on MBPP, Qwen 50 expertsexpert pruning · MBPP (Qwen)pass rate: 8.17 -> 9.34 (40 experts: 8.17 vs 8.17 tie)Mixed
remove 2 experts per layer with MoP (6 of 8)MoP · Mixtral inferencememory GB / speedup: Mem 87.87 -> 66.87, 1.22x speedupImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy