Skip to content
conftrace_

gradient stability

7 papers

Explore in graph

Co-occurring keywords · papers mentioning both

recurrent neural network (3) gradient descent (3) vanishing gradient (2) hidden state (2) style perturbation (1) hidden state permutation (1) low-curvature neural network (1) continuous-time attention (1) long-time dependency (1) gradient flossing (1)

Papers

Continuous-Time Attention: PDE-Guided Mechanisms for Long-Sequence Transformers EMNLP 2025 SVasP: Self-Versatility Adversarial Style Perturbation for Cross-Domain Few-Shot Learning AAAI 2025 Gradient Flossing: Improving Gradient Descent through Dynamic Control of Jacobians NeurIPS 2023 Efficient Training of Low-Curvature Neural Networks NeurIPS 2022 Orthogonal Graph Neural Networks AAAI 2022 Coupled Oscillatory Recurrent Neural Network (coRNN): An accurate and (gradient) stable architecture for learning long time dependencies ICLR 2021 Shuffling Recurrent Neural Networks AAAI 2021