Research map / Architectures and efficiency / RNNs and normalization
Temporal superposition and feature geometry of RNNs under memory demands
Pratyaksh Sharma, Alexandra Maria Proca, Lucas Prieto, Pedro A. M. Mediano
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 |
|---|---|---|---|
| increase temporal sparsity of inputs (ablation) | SSM on k-delay task · k-delay task (2D hidden) | angular spread of feature directions: approx 90 deg -> approx 270 deg | Mixed |
| use ReLU recurrence in addition to ReLU readout (ablation) | RNN on k-delay task (sparse regime) · k-delay task (2D hidden) | packing of intermediate features into interference-free space | Improved |
| add a nonlinear (ReLU) readout to linear RNN (ablation) | linear RNN · k-delay task, sparse inputs | projection interference | Improved |
| increase memory delay k with 5 spatial features (ablation) | nonlinear RNN, 2D hidden · k-delay task, 5 features | number of features represented: k=1: A,B,C kept; larger k: only A kept | Hurt |
Closest papers by meaning
- RNNs perform task computations by dynamically warping neural representations (NeurIPS 2025 · similarity 0.90)
- Geometry of naturalistic object representations in recurrent neural network models of working memory (NeurIPS 2024 · similarity 0.89)
- Dynamical phases of short-term memory mechanisms in RNNs (ICML 2025 · similarity 0.88)
- Task structure and nonlinearity jointly determine learned representational geometry (ICLR 2024 · similarity 0.88)
- Superposition Yields Robust Neural Scaling (NeurIPS 2025 · similarity 0.87)
- Learning dynamics in linear recurrent neural networks (ICML 2025 · similarity 0.87)
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.
