recurrence
Theoretical Foundations of Deep Selective State-Space Models
Similar ideas can be found in recent developments inspired by attention, such as RWKV [Peng et al.,
- Europe > United Kingdom > England > Oxfordshire > Oxford (0.04)
- Europe > Germany > Baden-Württemberg > Tübingen Region > Tübingen (0.04)
- Asia > China > Hong Kong (0.04)
- Europe > United Kingdom > England > Cambridgeshire > Cambridge (0.04)
Persistent Test-time Adaptation in Recurring Testing Scenarios
Current test-time adaptation (TT A) approaches aim to adapt a machine learning model to environments that change continuously.
- North America > United States > Illinois (0.04)
- Asia > Japan > Honshū > Kantō > Tokyo Metropolis Prefecture > Tokyo (0.04)
- North America > United States > New Jersey > Mercer County > Princeton (0.04)
- (3 more...)
- Information Technology (0.92)
- Health & Medicine (0.92)
- Asia > Middle East > Israel (0.04)
- North America > United States > California > Santa Clara County > Palo Alto (0.04)
- Europe > United Kingdom > England > Cambridgeshire > Cambridge (0.04)
- (3 more...)
Transformers Can Do Arithmetic with the Right Embeddings
We mend this problem by adding an embedding to each digit that encodes its position relative to the start of the number.
- Europe > Germany > Baden-Württemberg > Tübingen Region > Tübingen (0.04)
- South America > Chile > Santiago Metropolitan Region > Santiago Province > Santiago (0.04)
- North America > United States > Maryland (0.04)
- Europe > Italy > Tuscany > Florence (0.04)
- Research Report > Experimental Study (0.93)
- Research Report > New Finding (0.93)
- Government > Regional Government (0.46)
- Education (0.46)
- Banking & Finance (0.46)
- Information Technology > Artificial Intelligence > Natural Language > Large Language Model (1.00)
- Information Technology > Artificial Intelligence > Machine Learning > Neural Networks > Deep Learning (1.00)
- Information Technology > Artificial Intelligence > Natural Language > Chatbot (0.68)
- Information Technology > Artificial Intelligence > Representation & Reasoning (0.67)
Tight Dimension Independent Lower Bound on the Expected Convergence Rate for Diminishing Step Sizes in SGD
PHUONG_HA NGUYEN, Lam Nguyen, Marten van Dijk
Neural Information Processing Systems http://nips.cc/
- North America > United States > Connecticut (0.04)
- North America > Canada (0.04)
- Europe > United Kingdom > England > Cambridgeshire > Cambridge (0.04)
- Europe > Netherlands > South Holland > Dordrecht (0.04)
Hierarchically Gated Recurrent Neural Network for Sequence Modeling
Recently, there has been a renewed interest in using linear RNNs for efficient sequence modeling.
- North America > United States > Minnesota > Hennepin County > Minneapolis (0.14)
- North America > United States > Louisiana > Orleans Parish > New Orleans (0.04)
- North America > Canada > British Columbia > Vancouver (0.04)
- (11 more...)
Task-Driven Convolutional Recurrent Models of the Visual System
Aran Nayebi, Daniel Bear, Jonas Kubilius, Kohitij Kar, Surya Ganguli, David Sussillo, James J. DiCarlo, Daniel L. Yamins
Neural Information Processing Systems http://nips.cc/
- North America > United States > Massachusetts > Middlesex County > Cambridge (0.14)
- North America > United States > California > Santa Clara County > Stanford (0.05)
- Europe > Belgium > Flanders > Flemish Brabant > Leuven (0.04)
- (3 more...)
Brain-Like Object Recognition with High-Performing Shallow Recurrent ANNs
Jonas Kubilius, Martin Schrimpf, Kohitij Kar, Rishi Rajalingham, Ha Hong, Najib Majaj, Elias Issa, Pouya Bashivan, Jonathan Prescott-Roy, Kailyn Schmidt, Aran Nayebi, Daniel Bear, Daniel L. Yamins, James J. DiCarlo
Despite being significantly shallower than most models, CORnet-S is the top model on Brain-Score and outperforms similarly compact models on ImageNet.
- North America > Canada > British Columbia > Metro Vancouver Regional District > Vancouver (0.04)
- Europe > Belgium > Flanders > Flemish Brabant > Leuven (0.04)