MoRE: Mixture of Reused Experts
cs.LG, cs.AI
Submitted: 2026-09-16
Updated: 2026-09-17
Comments: Accepted to the Conference on Language Modeling (COLM 2026)
Code: https://github.com/EricQiu6/MoRE
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- Relaxed Recursive Transformers: Effective Parameter Sharing with Layer-wise LoRA
- The Devil is in the Detail: Simple Tricks Improve Systematic Generalization of Transformers
- DeepSeek-V3 Technical Report
- Do Generative Large Language Models need billions of parameters?
- Training Compute-Optimal Large Language Models
- Mixtral of Experts
- GShard: Scaling Giant Models with Conditional Computation and Automatic Sharding
- OLMoE: Open Mixture-of-Experts Language Models
- Olmo 3
- In-context Learning and Induction Heads
- Making Transformers Solve Compositional Tasks
- The FineWeb Datasets: Decanting the Web for the Finest Text Data at Scale
- Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layer
- Lessons on Parameter Sharing across Layers in Transformers
- ReXMoE: Reusing Experts with Minimal Overhead in Mixture-of-Experts
- Qwen3 Technical Report
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