Research map / Optimization / Implicit bias and generalization
Learning-to-Learn Stochastic Gradient Descent with Biased Regularization
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Use the online learned bias instead of zero bias (ITL) | Biased SGD · Synthetic regression with absolute loss | test performance | Helped |
| Use the online learned bias instead of zero bias (ITL) | Biased SGD · Synthetic classification with hinge loss | test performance | Helped |
| Use the online learned bias instead of zero bias (ITL) | Biased SGD · Lenk computer survey regression | test performance | Helped |
| Use the online learned bias instead of zero bias (ITL) | Biased SGD · Lenk computer survey classification | test performance | Helped |
| Process data in a single pass rather than using the more expensive ERM approach | Online biased SGD · Additional experiments | test performance | No clear effect |
Closest papers by meaning
- A Guide Through the Zoo of Biased SGD (NeurIPS 2023 · similarity 0.90)
- Direction Matters: On the Implicit Bias of Stochastic Gradient Descent with Moderate Learning Rate (ICLR 2021 · similarity 0.89)
- Implicit Regularization or Implicit Conditioning? Exact Risk Trajectories of SGD in High Dimensions (NeurIPS 2022 · similarity 0.88)
- Meta-learning with Stochastic Linear Bandits (ICML 2020 · similarity 0.88)
- SGD: The Role of Implicit Regularization, Batch-size and Multiple-epochs (NeurIPS 2021 · similarity 0.88)
- Non-asymptotic Analysis of Biased Adaptive Stochastic Approximation (NeurIPS 2024 · similarity 0.88)
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.