Research map / Architectures and efficiency / Quantization
Probabilistic Weight Fixing: Large-scale training of neural network weight uncertainties for quantisation.
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 PWFN instead of APoT | ResNet-18 · ImageNet | Top-1 accuracy: 69.9 -> 70.0 | Helped |
| use PWFN instead of APoT | ResNet-34 · ImageNet | Top-1 accuracy: 73.4 -> 74.3 | Helped |
| use PWFN instead of WFN | ResNet-50 · ImageNet | Top-1 accuracy: 76.0 -> 77.5 | Helped |
| use PWFN instead of Q-ViT | DeiT-Small · ImageNet | Entropy: 11.3 -> 2.7 | Helped |
| use PWFN instead of Q-ViT | DeiT-Tiny · ImageNet | Top-1 accuracy: 69.6 -> 71.2 | Helped |
| add relative-distance prior initialization (ablation) | PWFN on ResNet-18 · ImageNet | Top-1 accuracy: 69.3 (69.6) -> 70.0 (70.1) | Helped |
| add relative-distance prior initialization (ablation) | PWFN on ResNet-34 · ImageNet | Top-1 accuracy: 73.5 (74.4) -> 74.3 (74.6) | Helped |
| add relative-distance prior initialization (ablation) | PWFN on ResNet-50 · ImageNet | Top-1 accuracy: 77.2 (78.1) -> 77.5 (78.3) | Helped |
| add relative-distance prior initialization (ablation) | PWFN on DeiT-Small · ImageNet | Top-1 accuracy: 78.0 (78.3) -> 78.1 (78.5) | Helped |
| add relative-distance prior initialization (ablation) | PWFN on DeiT-Tiny · ImageNet | Top-1 accuracy: 71.4 (71.6) -> 71.2 (71.5) | Hurt |
Closest papers by meaning
- Sparse Uncertainty Representation in Deep Learning with Inducing Weights (NeurIPS 2021 · similarity 0.88)
- Efficient and Scalable Bayesian Neural Nets with Rank-1 Factors (ICML 2020 · similarity 0.87)
- A Symmetry-Aware Exploration of Bayesian Neural Network Posteriors (ICLR 2024 · similarity 0.86)
- Make Me a BNN: A Simple Strategy for Estimating Bayesian Uncertainty from Pre-trained Models (CVPR 2024 · similarity 0.86)
- Training Bayesian Neural Networks with Sparse Subspace Variational Inference (ICLR 2024 · similarity 0.86)
- A Study of Bayesian Neural Network Surrogates for Bayesian Optimization (ICLR 2024 · 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.