Beyond the Previous Layer: Residual Predictive Structure in Sparse MoE Routing

arXiv:2609.17940 · cs.LG · Submitted 2026-09-15 · Read on arXiv

cs.LG

Submitted: 2026-09-15

Updated: 2026-09-15

Comments: 6 pages, 3 tables

Code: https://github.com/withfanta/moe-routing-dynamics

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

Terminology

• Provenance — 6× in this paper · explained in 3 episodes
• Linear probe — 3× in this paper · explained in 2 episodes
• Reinforcement Learning — 2× in this paper · explained in 4 episodes
• Actionability — 1× in this paper · explained in 3 episodes
• Decode — 1× in this paper · explained in 2 episodes
• Mixture of Experts — 1× in this paper · explained in 2 episodes
• Mixture-of-Experts (MoE) — 1× in this paper · explained in 7 episodes

Sources

• Layerwise Recurrent Router for Mixture-of-Experts
• Path-Constrained Mixture-of-Experts
• Evidence for Shared Routing Geometry and Dynamics in Sparse Mixture-of-Experts
• Expert-Space Exploration in MoE Reinforcement Learning
• Do Dynamic Routers Need Memory? HeRo: History-Aware Routing for Efficient LLM Inference
• OLMoE: Open Mixture-of-Experts Language Models
• Pointer Sentinel Mixture Models

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