Research map / Label-efficient and robust learning / Recognition and long-tail
HCL-FF: Hierarchical and Contrastive Learning for Forward-Forward Algorithm
Jie-En Yao, Hong-En Chen, C.-C. Jay Kuo
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 |
|---|---|---|---|
| add contrastive objective (V2 to V3) (ablation) | HCL-FF · CIFAR-100 | accuracy: 54.65 -> 63.77 | Improved |
| add contrastive objective (V2 to V3) (ablation) | HCL-FF · CIFAR-10 | accuracy: 88.80 -> 91.05 | Improved |
| add hierarchical objective (V2 to V4) (ablation) | HCL-FF · CIFAR-100 | accuracy: 54.65 -> 61.61 | Improved |
| add hierarchical objective (V2 to V4) (ablation) | HCL-FF · CIFAR-10 | accuracy: 88.80 -> 88.20 | Hurt |
| add strong data augmentation without hierarchy (V1 to V2) (ablation) | HCL-FF baseline · CIFAR-10 | accuracy: 86.22 -> 88.80 | Improved |
| add strong augmentation with hierarchy and contrastive (V5 to V6) (ablation) | HCL-FF · CIFAR-100 | accuracy: 65.22 -> 70.76 | Improved |
| add residual shortcuts (ablation) | HCL-FF · CIFAR-100 | accuracy: 65.43 -> 70.76 | Improved |
| add residual shortcuts (ablation) | HCL-FF · CIFAR-10 | accuracy: 90.85 -> 91.83 | Improved |
| use WordNet hierarchy instead of data-driven (ablation) | HCL-FF · CIFAR-100 | accuracy: 70.76 -> 71.01 | No clear effect |
| use Word2Vec hierarchy instead of data-driven (ablation) | HCL-FF · CIFAR-100 | accuracy: 70.76 -> 69.59 | Hurt |
| use balanced mapping instead of decremental (ablation) | HCL-FF · CIFAR-100 | accuracy: 68.40 -> 70.76 | Improved |
| use selective interval prediction instead of last layer (ablation) | HCL-FF · CIFAR-100 | accuracy: 67.42 -> 70.76 | Improved |
Closest papers by meaning
- Bridging Mini-Batch and Asymptotic Analysis in Contrastive Learning: From InfoNCE to Kernel-Based Losses (ICML 2024 · similarity 0.86)
- FFF: Fixing Flawed Foundations in Contrastive Pre-Training Results in Very Strong Vision-Language Models (CVPR 2024 · similarity 0.86)
- Convolutional Channel-Wise Competitive Learning for the Forward-Forward Algorithm (AAAI 2024 · similarity 0.85)
- BECLR: Batch Enhanced Contrastive Few-Shot Learning (ICLR 2024 · similarity 0.85)
- Variational Supervised Contrastive Learning (NeurIPS 2025 · similarity 0.84)
- Hierarchical Selective Classification (NeurIPS 2024 · similarity 0.84)
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.
