AI Paper Atlas

Research map / Architectures and efficiency / Mixture of experts

Unveiling Super Experts in Mixture-of-Experts Large Language Models

Zunhai Su, Qingyuan Li, HaoZhang, Weihao Ye, Qibo Xue, Yulei Qian, Ngai Wong, Kehong Yuan

ICLR 2026 · Architectures and efficiency › Mixture of experts · approach: Expert routing and specialization

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
prune the Super Experts (0.05-0.39% of experts) (ablation)Qwen3-30B-A3B, DeepSeek-V2-Lite, Mixtral-8x7B · 9 non-reasoning benchmarksaverage accuracy: 70.22 -> 55.00 (Qwen3-30B-A3B); 60.27 -> 43.90 (DeepSeek-V2-Lite); 67.84 -> 49.38 (Mixtral)Hurt
prune random experts of equal number (global or layer-controlled) (ablation)Qwen3-30B-A3B, DeepSeek-V2-Lite, Mixtral-8x7B · 9 non-reasoning benchmarksaverage accuracy: 70.22 -> 70.36 (random), 70.21 (LC random) on Qwen3-30B-A3BNo clear effect
prune the SEs (ablation)Qwen3-30B-A3B · GSM8Kaccuracy: 89.61 -> 42.38Hurt
prune the SEs (ablation)DeepSeek-R1 · reasoning benchmarks (GPQA, Math-500, AIME, LiveCodeBench)average Pass@1: 75.64 -> 1.81Hurt
prune the SEs (ablation)Qwen3-30B-A3B (thinking) · reasoning benchmarks (GPQA, Math-500, AIME, HumanEval)average Pass@1: 69.37 -> 4.02Hurt
prune the SEs (ablation)Qwen3-30B-A3B · WikiText-2perplexity: 8.70 -> 59.86Hurt
prune final-layer outlier experts instead of SEs (ablation)Qwen3-30B-A3B · WikiText-2perplexity: 8.70 -> 8.71No clear effect
prune final-layer outlier experts instead of SEs (ablation)DeepSeek-V2-Lite · WikiText-2perplexity: 6.31 -> 6.32No clear effect
prune SEs to remove massive activations (ablation)Qwen3-30B-A3B · attention sinks on C4attention sink decay rate: approximately or even exceeding 90%Hurt
change SE identification threshold (P95 to P99.9, 0.07 to 0.1 of max) (ablation)SE profiling · Qwen3-30B-A3B, DeepSeek-V2-Lite, Mixtral on C4 and WikiText-2number of identified SEs: 3, 2 and 1 SEs identified at all thresholdsNo clear effect
switch dataset used for SE profiling (C4, WikiText-2, C-Eval, GSM8K, HumanEval, OoD) (ablation)SE profiling · Qwen3-30B-A3B and othersSE distributionNo clear effect
compare base vs post-trained checkpoints (Base, Instruct-2507, Thinking-2507) (ablation)Qwen3-30B-A3B · SE locationsSE setNo clear effect

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