Research map / Optimization / Optimizers
Per-example Gradients: a New Frontier for Understanding and Improving Optimizers
Vincent Roulet, Atish Agarwala
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| place sign last (SignEMA) instead of before averaging (MicroSignSGD) (ablation) | signSGD variants · C4 (151M decoder LM, B=64) | training loss | Improved |
| place sign after averaging but before EMA (SignSGD) instead of per-example sign (ablation) | signSGD variants · C4 (151M decoder LM, B=64) | training loss | Improved |
| use per-example mean of squared gradients as preconditioner (MicroAdam) | Adam · C4 (151M decoder LM, B=64) | eval loss: 3.06 vs 3.10 | Hurt |
| use variance-only preconditioner (MicroAdamVar) (ablation) | MicroAdam · C4 (151M decoder LM, B=64) | training loss | Hurt |
| use ReLU-filtered mean-squared preconditioner with clipping (MicroAdamMSQ) | Adam · C4 (151M decoder LM, B=64) | eval loss: 3.06 -> 3.05 | Improved |
| add gradient clipping to MicroAdam (ablation) | MicroAdam · C4 (151M decoder LM, B=64) | training spikes | Improved |
| use unfiltered mean-squared estimator without ReLU (ablation) | MicroAdamMSQ · C4 (151M decoder LM, B=64) | stability | Hurt |
| use per-example mean-squared preconditioner (MicroAdamMSQ) for compute | Adam · C4 (151M decoder LM, B=64) | wall-clock time: 20% more wall-clock time | Hurt |
Closest papers by meaning
- Coherent Gradients: An Approach to Understanding Generalization in Gradient Descent-based Optimization (ICLR 2020 · similarity 0.88)
- Adam-mini: Use Fewer Learning Rates To Gain More (ICLR 2025 · similarity 0.87)
- On The Concurrence of Layer-wise Preconditioning Methods and Provable Feature Learning (ICML 2025 · similarity 0.86)
- Understanding the Generalization of Stochastic Gradient Adam in Learning Neural Networks (NeurIPS 2025 · similarity 0.86)
- Surge Phenomenon in Optimal Learning Rate and Batch Size Scaling (NeurIPS 2024 · similarity 0.86)
- Small Batch Size Training for Language Models: When Vanilla SGD Works, and Why Gradient Accumulation is Wasteful (NeurIPS 2025 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
