Research map / Architectures and efficiency / Quantization
Multi-Precision Quantized Neural Networks via Encoding Decomposition of {-1,+1}
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 encoding precision from 1-bit to 8-bit for activations and weights (ablation) | MBN on CIFAR-10 · CIFAR-10 | accuracy: 90.39% -> 90.43% | Mixed |
| increase encoding precision from 1-bit to 8-bit for activations and weights (ablation) | MBN on ImageNet · ImageNet | Top-1 accuracy: 47.10% -> 68.63% | Helped |
| increase encoding precision from 1-bit to 8-bit for activations and weights (ablation) | MBN on ImageNet · ImageNet | Top-5 accuracy: 71.70% -> 88.70% | Helped |
| increase encoding precision from 8-bit to 6-bit for activations and weights (ablation) | MBN with SSD · Pascal VOC2007 test | mAP: 0.6351 -> 0.6131 | Hurt |
| reduce encoding precision from 6-bit to 5-bit for activations and weights (ablation) | MBN with SSD · Pascal VOC2007 test | mAP: 0.6131 -> 0.5423 | Hurt |
| use 8-bit encoding instead of full precision | MBN with SSD · Pascal VOC2007 test | mAP: 0.6392 -> 0.6351 | No clear effect |
Closest papers by meaning
- Adaptive Loss-Aware Quantization for Multi-Bit Networks (CVPR 2020 · similarity 0.90)
- One-Shot Model for Mixed-Precision Quantization (CVPR 2023 · similarity 0.88)
- Distribution-Aware Adaptive Multi-Bit Quantization (CVPR 2021 · similarity 0.88)
- Two-Step Quantization for Low-Bit Neural Networks (CVPR 2018 · similarity 0.88)
- Scalable methods for 8-bit training of neural networks (NeurIPS 2018 · similarity 0.87)
- QBB: Quantization with Binary Bases for LLMs (NeurIPS 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.