You Only Need 2/3 of the Chosen Experts: An Empirical Study of Dynamic Expert Pruning in Fine-Grained MoE LLMs

arXiv:2609.25809 · cs.LG, cs.AI · Submitted 2026-09-22 · Read on arXiv

cs.LG, cs.AI

Submitted: 2026-09-22

Updated: 2026-09-22

Comments: 25 pages, 4 figures

Code: https://github.com/huggingface/lighteval

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Pruning — 81× in this paper · explained in 4 episodes
• Redundancy — 19× in this paper · explained in 2 episodes
• Benchmark — 14× in this paper · explained in 2 episodes
• Calibration — 11× in this paper · explained in 7 episodes
• Core — 9× in this paper · explained in 2 episodes
• Decode — 8× in this paper · explained in 2 episodes
• Accuracy — 5× in this paper · explained in 1 episode
• Transformer — 5× in this paper · explained in 2 episodes
• Adaptation — 4× in this paper · explained in 1 episode
• Mixture-of-Experts (MoE) — 2× in this paper · explained in 7 episodes
• Multimodal Dataset — 1× in this paper · explained in 2 episodes
• Pairwise Comparisons — 1× in this paper · explained in 2 episodes
• Regularization — 1× in this paper · explained in 3 episodes
• Sparsity — 1× in this paper · explained in 4 episodes
• Verifier — 1× in this paper · explained in 2 episodes

Sources

• Qwen3-VL Technical Report
• PIQA: Reasoning about Physical Commonsense in Natural Language
• Are We on the Right Way for Evaluating Large Vision-Language Models?
• Ban&Pick: Enhancing Performance and Efficiency of MoE-LLMs via Smarter Routing
• BoolQ: Exploring the Surprising Difficulty of Natural Yes/No Questions
• Think you have Solved Question Answering? Try ARC, the AI2 Reasoning Challenge
• Training Verifiers to Solve Math Word Problems
• DeepSeek-V2: A Strong, Economical, and Efficient Mixture-of-Experts Language Model
• DeepSeek-V4: Towards Highly Efficient Million-Token Context Intelligence
• Gemma 4 Technical Report
• GLM-5: from Vibe Coding to Agentic Engineering
• VizWiz Grand Challenge: Answering Visual Questions from Blind People
• Measuring Massive Multitask Language Understanding
• Measuring Mathematical Problem Solving With the MATH Dataset
• GQA: A New Dataset for Real-World Visual Reasoning and Compositional Question Answering
• LiveCodeBench: Holistic and Contamination Free Evaluation of Large Language Models for Code
• Mixtral of Experts
• Kimi K3: Open Frontier Intelligence
• Efficient Memory Management for Large Language Model Serving with PagedAttention
• Let's Verify Step by Step

Related papers

• Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
• AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
• Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
• Convergence issues in Relational Concept Analysis based on AOC-posets
• Beliefs Beyond Posteriors: Local-Consistency Optimisation for Bayesian Neural Networks
• Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks