Tensorion: A Tensor-Aware Generalization of the Muon Optimizer
cs.LG, cs.CV, cs.NA, math.NA
Submitted: 2026-06-24
Updated: 2026-08-27
Code: https://github.com/MTML-LAB/Tensorion
Terminology
Sources
- Old Optimizer, New Norm: An Anthology
- Muon Optimizes Under Spectral Norm Constraints
- An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
- Dion: Distributed Orthonormalized Updates
- Muon is Scalable for LLM Training
- Decoupled Weight Decay Regularization
- Most tensor problems are NP-hard
- A New Perspective on Shampoo's Preconditioner
- Adam: A Method for Stochastic Optimization
- Kimi K2: Open Agentic Intelligence
- SOAP: Improving and Stabilizing Shampoo using Adam
- Gluon: Making Muon & Scion Great Again! (Bridging Theory and Practice of LMO-based Optimizers for LLMs)
- Fantastic Four: Differentiable Bounds on Singular Values of Convolution Layers
- TEON: Tensorized Orthonormalization Beyond Layer-Wise Muon for Large Language Model Pre-Training
- Concentration Inequalities for Sub-Weibull Random Tensors
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