Research map / Architectures and efficiency / RNNs and normalization
Push-pull Feedback Implements Hierarchical Information Retrieval Efficiently
Xiao Liu, Xiaolong Zou, Zilong Ji, Gengshuo Tian, Yuanyuan Mi, Tiejun Huang, K. Y. Michael Wong, Si Wu
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 sequential push then pull feedback between layers | two-layer hierarchical Hopfield-type network · VGG-derived cat/dog image memory patterns, 1800 images | layer-1 retrieval accuracy: 71.04% | Improved |
| apply push (positive) feedback early (ablation) | two-layer hierarchical network · cat/dog image patterns | retrieval accuracy of cousin (other-category) patterns | Improved |
| apply pull (negative) feedback later (ablation) | two-layer hierarchical network · cat/dog image patterns | retrieval accuracy of target child pattern | Improved |
| add pull feedback W = -b1 delta to subtract part of the parent mean (ablation) | two-layer hierarchical network · synthetic hierarchical patterns, N=2000 | intra-class noise tail causing retrieval errors | Improved |
| use discrete instead of continuous Hopfield dynamics (ablation) | push-pull feedback · synthetic hierarchical patterns | overall joint effect of push-pull | Hurt |
Closest papers by meaning
- Feedback control guides credit assignment in recurrent neural networks (NeurIPS 2024 · similarity 0.82)
- Factual recall in linear associative memories: sharp asymptotics and mechanistic insights (NeurIPS 2026 · similarity 0.82)
- Forward Learning with Top-Down Feedback: Empirical and Analytical Characterization (ICLR 2024 · similarity 0.81)
- NeuroPath: Neurobiology-Inspired Path Tracking and Reflection for Semantically Coherent Retrieval (NeurIPS 2025 · similarity 0.81)
- Efficient Target Propagation by Deriving Analytical Solution (AAAI 2024 · similarity 0.81)
- Dynamical phases of short-term memory mechanisms in RNNs (ICML 2025 · similarity 0.81)
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.
