Research map / Optimization / Optimizers
Stochastic Multiple Target Sampling Gradient Descent
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 |
|---|---|---|---|
| increase number of particles from 3 to 5 (ablation) | MT-SGD · Multi-MNIST/Fashion/Fashion+MNIST | accuracy | Helped |
| increase number of particles beyond 5 (7, 9, 15) (ablation) | MT-SGD · Multi-MNIST/Fashion/Fashion+MNIST | accuracy | No clear effect |
| increase particles from 3 to 5 (ablation) | MT-SGD · Multi-MNIST | ECE | Helped |
| increase particles to 7, 9, 15 (ablation) | MT-SGD · Multi-MNIST | ECE | No clear effect |
| use MT-SGD instead of MOO-SVGD | MT-SGD · SARCOS validation | MSE: 0.2459 -> 0.2251 | Helped |
| use MT-SGD instead of MOO-SVGD | MT-SGD · SARCOS test | MSE: 0.0515 -> 0.0428 | Helped |
| use MT-SGD instead of MOO-SVGD | MT-SGD · CelebA first 10 tasks | ECE (%): 2.5 -> 2.0 | Helped |
| use MT-SGD instead of Single task | MT-SGD · CelebA first 10 tasks | accuracy (%): 88.3 -> 89.0 | Helped |
| use MT-SGD instead of MOO-SVGD on BL task | MT-SGD · CelebA BL attribute | accuracy (%): 86.4 -> 86.3 | Hurt |
| use MT-SGD instead of MGDA | MT-SGD · CelebA 40 attributes | error (%): 8.73 -> 8.17 | Helped |
| use MT-SGD instead of Pareto MTL | MT-SGD · Multi-Fashion+MNIST top left | ECE (%): 9.44 -> 4.65 | Helped |
| use MT-SGD instead of second-best baseline | MT-SGD · Multi-task image benchmarks | accuracy | Helped |
Closest papers by meaning
- Towards Understanding the Dynamics of Gaussian-Stein Variational Gradient Descent (NeurIPS 2023 · similarity 0.89)
- Stein Variational Gradient Descent as Moment Matching (NeurIPS 2018 · similarity 0.89)
- A Convergence Theory for SVGD in the Population Limit under Talagrand’s Inequality T1 (ICML 2022 · similarity 0.89)
- Message Passing Stein Variational Gradient Descent (ICML 2018 · similarity 0.89)
- Stein Variational Gradient Descent With Matrix-Valued Kernels (NeurIPS 2019 · similarity 0.89)
- ELBOing Stein: Variational Bayes with Stein Mixture Inference (ICLR 2025 · 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.