Latest Articles for Computational Models

Averaging enhances KANs' performance and stability in machine learning tasks.

Exploring how weight matrices evolve during machine learning training.

Exploring geometrical perspectives on density functional theory in spin-lattice models.

This study reveals how neural networks can accurately remember and reproduce spike patterns under varied conditions.

A study on the behavior of cellular automata and random networks regarding memory tasks.

A look at neural networks using detailed and simplified models.

GATH enhances knowledge graph completion using advanced attention mechanisms for better accuracy.

A new model improves sequence generation by combining the strengths of VAEs and SSMs.

Logifolds improve understanding and accuracy in analyzing complex datasets.

A new approach using Kolmogorov-Arnold Networks improves predictions of nuclear binding energy.

Innovations in adaptive LIF neurons enhance performance in temporal and spatial tasks.

Examining how transformers learn from context without needing retraining.

New modeling techniques enhance our understanding of bacterial movement.

Research reveals unique symmetry restoration in many-body localized systems through the Mpemba effect.

Exploring how defects influence entanglement in quantum systems.

A new method using AutoSparse for efficient neural network pruning at the start.

New methods improve predictions of coastal dynamics using deep learning.

A look into Sparse Mamba, a method for better language model control.

A novel approach to enhance GFlowNet training with policy-dependent rewards.

This article presents a model to study error spread in quantum computing.

This article discusses the connection between RvNNs and Transformers through CRvNN and NDR.

A new method enhances language understanding in Transformer models using non-linear geometries.

Exploring how errors in catalytic computation can expand computational capabilities.

Explore the complexity of finite automata using translucent letters in language recognition.

HetSheaf improves data representation in heterogeneous graphs for better model performance.

An overview of string functions and their significance in computing.

Exploring the impact of recurrence on the effectiveness of Transformers in language tasks.

Utilizing reduced-order modeling for faster earthquake ground motion predictions.

New methods enhance GNNs for challenging graph types.

Examining how BERT interprets words with multiple meanings.

Researchers develop models that mimic brain processing using light and superconducting systems.

A new method enhances understanding and reliability of neural networks.

A new method improves chemical structure representation for enhanced analysis and efficiency.

Quantum systems offer a reliable method for producing unpredictable numbers.

This article discusses methods to better understand neural networks through Sparse Autoencoders and Mutual Feature Regularization.

Exploring the capabilities and challenges of Transformer technology in understanding language.

A fresh method to understand causal relationships in dynamic environments.

This research improves Knowledge Graphs using refined negative sampling techniques for better model performance.

Discover how NERDSS models particle interactions and reveals complex patterns in nature.

Exploring how different neurons enhance brain performance and influence machine learning.