# On Surprising Effectiveness of Masking Updates in Adaptive Optimizers

arXiv: https://arxiv.org/abs/2602.15322 · HTML: https://mlsanity.com/paper/2602.15322/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- Builds on 55 of its 69 references that are in the map.
- First posted on arXiv 17 Feb 2026.

## Builds on (most cited first, 30 of 55)

- [Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://mlsanity.com/paper/1910.10683/) (2019)
- [PaLM: Scaling Language Modeling with Pathways](https://mlsanity.com/paper/2204.02311/) (2022)
- [Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layer](https://mlsanity.com/paper/1701.06538/) (2017)
- [Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity](https://mlsanity.com/paper/2101.03961/) (2021)
- [The Pile: An 800GB Dataset of Diverse Text for Language Modeling](https://mlsanity.com/paper/2101.00027/) (2020)
- [GShard: Scaling Giant Models with Conditional Computation and Automatic Sharding](https://mlsanity.com/paper/2006.16668/) (2020)
- [Sharpness-Aware Minimization for Efficiently Improving Generalization](https://mlsanity.com/paper/2010.01412/) (2020)
- [On Large-Batch Training for Deep Learning: Generalization Gap and Sharp Minima](https://mlsanity.com/paper/1609.04836/) (2016)
- [Optimization Methods for Large-Scale Machine Learning](https://mlsanity.com/paper/1606.04838/) (2016)
- [On Layer Normalization in the Transformer Architecture](https://mlsanity.com/paper/2002.04745/) (2020)
- [GLaM: Efficient Scaling of Language Models with Mixture-of-Experts](https://mlsanity.com/paper/2112.06905/) (2021)
- [Optimizing Neural Networks with Kronecker-factored Approximate Curvature](https://mlsanity.com/paper/1503.05671/) (2015)
- [Transformers learn in-context by gradient descent](https://mlsanity.com/paper/2212.07677/) (2022)
- [What Can Transformers Learn In-Context? A Case Study of Simple Function Classes](https://mlsanity.com/paper/2208.01066/) (2022)
- [What learning algorithm is in-context learning? Investigations with linear models](https://mlsanity.com/paper/2211.15661/) (2022)
- [GaLore: Memory-Efficient LLM Training by Gradient Low-Rank Projection](https://mlsanity.com/paper/2403.03507/) (2024)
- [Shampoo: Preconditioned Stochastic Tensor Optimization](https://mlsanity.com/paper/1802.09568/) (2018)
- [8-bit Optimizers via Block-wise Quantization](https://mlsanity.com/paper/2110.02861/) (2021)
- [ST-MoE: Designing Stable and Transferable Sparse Expert Models](https://mlsanity.com/paper/2202.08906/) (2022)
- [Why gradient clipping accelerates training: A theoretical justification for adaptivity](https://mlsanity.com/paper/1905.11881/) (2019)
- [ASAM: Adaptive Sharpness-Aware Minimization for Scale-Invariant Learning of Deep Neural Networks](https://mlsanity.com/paper/2102.11600/) (2021)
- [Identifying and attacking the saddle point problem in high-dimensional non-convex optimization](https://mlsanity.com/paper/1406.2572/) (2014)
- [Three Factors Influencing Minima in SGD](https://mlsanity.com/paper/1711.04623/) (2017)
- [Coordinate Descent Algorithms](https://mlsanity.com/paper/1502.04759/) (2015)
- [ADAHESSIAN: An Adaptive Second Order Optimizer for Machine Learning](https://mlsanity.com/paper/2006.00719/) (2020)
- [SOAP: Improving and Stabilizing Shampoo using Adam](https://mlsanity.com/paper/2409.11321/) (2024)
- [DeepNet: Scaling Transformers to 1,000 Layers](https://mlsanity.com/paper/2203.00555/) (2022)
- [Transformers without Tears: Improving the Normalization of Self-Attention](https://mlsanity.com/paper/1910.05895/) (2019)
- [Adding Gradient Noise Improves Learning for Very Deep Networks](https://mlsanity.com/paper/1511.06807/) (2015)
- [Why Transformers Need Adam: A Hessian Perspective](https://mlsanity.com/paper/2402.16788/) (2024)

## Built on by (most cited first, 7 of 7)

- [Dion3: Full-Stack Orthogonal Updates](https://mlsanity.com/paper/2608.11612/) (2026)
- [Hierarchical Muon: Tiled Newton-Schulz Updates for Efficient Muon Optimization](https://mlsanity.com/paper/2606.27216/) (2026)
- [FedACT: Federated Adaptive Coordinate Trust Modulation for Robust Transformer Training under Data Heterogeneity](https://mlsanity.com/paper/2607.03763/) (2026)
- [OmniOpt: Taxonomy, Geometry, and Benchmarking of Modern Optimizers](https://mlsanity.com/paper/2607.04033/) (2026)
- [Dropout and Random Gradient Masking Are Asymptotically Equivalent in Large ResNets](https://mlsanity.com/paper/2607.16761/) (2026)
- [How Much Regularization Survives Averaging? Update Masking in Federated Learning](https://mlsanity.com/paper/2608.23286/) (2026)
- [ExpertMuon-Compass: Alignment-Guided Step Sizes for Mixture-of-Experts Training](https://mlsanity.com/paper/2610.04140/) (2026)

## Abstract (arXiv)

Training large language models (LLMs) relies almost exclusively on dense adaptive optimizers with increasingly sophisticated preconditioners. We challenge this by showing that randomly masking parameter updates can be highly effective, with a masked variant of RMSProp consistently outperforming recent state-of-the-art optimizers. Our analysis reveals that the random masking induces a curvature-dependent geometric regularization that smooths the optimization trajectory. Motivated by this finding, we introduce Momentum-aligned gradient masking (Magma), which modulates the masked updates using momentum-gradient alignment. Extensive LLM pre-training experiments show that Magma is a simple drop-in replacement for adaptive optimizers with consistent gains and negligible computational overhead. Notably, for the 1B model size, Magma reduces perplexity by over 19% and 9% compared to Adam and Muon, respectively.