Research map / Architectures and efficiency / RNNs and normalization
PACEAttention: Principled and Adaptive Feature Compression-Expansion Grounded in the Geometry of $\text{MCR}^2$
Xiaojie Yu, Haibo Zhang, Jeremiah D. Deng, Lizhi Peng
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 softplus non-negativity constraint on expansion weights (ablation) | PACENet-B · CIFAR-10 | accuracy vs r | Improved |
| keep MLP in ToST (vs removing it) (ablation) | ToST-S · CIFAR-10 | accuracy: 81.14 -> 84.50 | Improved |
| use PACEAttention (PACENet-S, 6M, r=10) instead of ToST-S without MLP | ToST-S-w/o MLP · CIFAR-10 | accuracy: 81.14 -> 81.67 | Improved |
| increase r from 10 to 40 (ablation) | PACENet-S (6M) · CIFAR-10 | accuracy: 81.67 -> 82.17 | Improved |
| scale from PACENet-S (6M) to PACENet-B (20M) at r=40 (ablation) | PACENet · CIFAR-10 | accuracy: 82.17 -> 83.85 | Improved |
| increase r beyond 20 (ablation) | PACENet · CIFAR-10 | accuracy | No clear effect |
| scale PACENet from S to B+ | PACENet · ImageNet ReaL | top-1 accuracy: 79.0 -> 84.6 | Improved |
Closest papers by meaning
- A Global Geometric Analysis of Maximal Coding Rate Reduction (ICML 2024 · similarity 0.85)
- Distribution-Aware Tensor Decomposition for Compression of Convolutional Neural Networks (NeurIPS 2025 · similarity 0.84)
- The Feature Speed Formula: a flexible approach to scale hyper-parameters of deep neural networks (NeurIPS 2024 · similarity 0.83)
- Compress to Impress: Efficient LLM Adaptation Using a Single Gradient Step on 100 Samples (NeurIPS 2025 · similarity 0.83)
- Beyond Matryoshka: Revisiting Sparse Coding for Adaptive Representation (ICML 2025 · similarity 0.83)
- SMMF: Square-Matricized Momentum Factorization for Memory-Efficient Optimization (AAAI 2025 · similarity 0.83)
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.
