arXiv subjects

M. Bilal

Publications and source records attributed to M. Bilal.

3 recordsLinked to original sources

arXiv2026

Scalable Machines with Intrinsic Higher Mental-State Dynamics

Drawing on recent breakthroughs in cellular neurobiology and detailed biophysical modeling linking neocortical pyramidal neurons to distinct mental-state regimes, this work introduces a mathematically grounded formulation showing how models (e.g., Transformers) can implement computational principles underlying awake imaginative thought to pre-select relevant information before attention is applied via triadic modulation loops among queries ($Q$), keys ($K$), and values ($V$).~Scalability experiments on ImageNet-1K, benchmarked against a standard Vision Transformer (ViT), demonstrate significantly faster learning with reduced computational demand (fewer heads, layers, and tokens), consistent with our prior findings in reinforcement learning and language modeling. The approach operates at approximately $\mathcal{O}(N)$ complexity with respect to the number of input tokens $N$.

cs.LG↗

arXiv2020

Blended Multi-Modal Deep ConvNet Features for Diabetic Retinopathy Severity Prediction

Diabetic Retinopathy (DR) is one of the major causes of visual impairment and blindness across the world. It is usually found in patients who suffer from diabetes for a long period. The major focus of this work is to derive optimal representation of retinal images that further helps to improve the performance of DR recognition models. To extract optimal representation, features extracted from multiple pre-trained ConvNet models are blended using proposed multi-modal fusion module. These final representations are used to train a Deep Neural Network (DNN) used for DR identification and severity level prediction. As each ConvNet extracts different features, fusing them using 1D pooling and cross pooling leads to better representation than using features extracted from a single ConvNet. Experimental studies on benchmark Kaggle APTOS 2019 contest dataset reveals that the model trained on proposed blended feature representations is superior to the existing methods. In addition, we notice that cross average pooling based fusion of features from Xception and VGG16 is the most appropriate for DR recognition. With the proposed model, we achieve an accuracy of 97.41%, and a kappa statistic of 94.82 for DR identification and an accuracy of 81.7% and a kappa statistic of 71.1% for severity level prediction. Another interesting observation is that DNN with dropout at input layer converges more quickly when trained using blended features, compared to the same model trained using uni-modal deep features.

eess.IV↗

arXiv2019

QED radiative corrections to the $^2$P$_{1/2}$-$^2$P$_{3/2}$ fine-structure in fluorinelike ions

Ab initio calculations of QED radiative corrections to the $^2P_{1/2}$ - $^2P_{3/2}$ fine-structure transition energy are performed for selected F-like ions. These calculations are nonperturbative in $αZ$ and include all first-order and many-electron second-order effects in $α$. When compared to approximate QED computations, a notable discrepancy is found especially for F-like uranium for which the predicted self-energy contributions even differ in sign. Moreover, all deviations between theory and experiment for the $^2P_{1/2}$ - $^2P_{3/2}$ fine-structure energies of F-like ions, reported recently by Li et al., Phys. Rev. A 98, 020502(R) (2018), are resolved if their highly accurate, non-QED fine-structure values are combined with the QED corrections ab initially evaluated here.

physics.atom-ph↗