arXiv subjects

Karanpartap Singh

Publications and source records attributed to Karanpartap Singh.

5 recordsLinked to original sources

arXiv2026

Interpretable Cross-Network Attention for Resting-State fMRI Representation Learning

Understanding how large-scale functional brain networks reorganize during cognitive decline remains a central challenge in neuroimaging. While recent self-supervised models have shown promise for learning representations from resting-state fMRI, their internal mechanisms are difficult to interpret, limiting mechanistic insight. We propose BrainInterNet, a network-aware self-supervised framework based on masked reconstruction with cross-attention that explicitly models inter-network dependencies in rs-fMRI. By selectively masking predefined functional networks and reconstructing them from remaining context, our approach enables direct quantification of network predictability and interpretable analysis of cross-network interactions. We train BrainInterNet on multi-cohort fMRI data (from the ABCD, HCP Development, HCP Young Adults, and HCP Aging datasets) and evaluate on the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset, in total comprising 5,582 recordings. Our method reveals systematic alterations in the brain's network interactions under AD, including in the default mode, limbic, and attention networks. In parallel, the learned representations support accurate Alzheimer's-spectrum classification and yield a compact summary marker that tracks disease severity longitudinally. Together, these results demonstrate that network-guided masked modeling with cross-attention provides an interpretable and effective framework for characterizing functional reorganization in neurodegeneration.

cs.LG↗

arXiv2025

Curriculum-Guided Layer Scaling for Language Model Pretraining

As the cost of pretraining large language models grows, there is continued interest in strategies to improve learning efficiency during this core training stage. Motivated by cognitive development, where humans gradually build knowledge as their brains mature, we propose Curriculum-Guided Layer Scaling (CGLS), a framework for compute-efficient pretraining that synchronizes increasing data difficulty with model growth through progressive layer stacking (i.e. gradually adding layers during training). At the 100M parameter scale, using a curriculum transitioning from synthetic short stories to general web data, CGLS outperforms baseline methods on the question-answering benchmarks PIQA and ARC. Pretraining at the 1.2B scale, we stratify the DataComp-LM corpus with a DistilBERT-based classifier and progress from general text to highly technical or specialized content. Our results show that progressively increasing model depth alongside sample difficulty leads to better generalization and zero-shot performance on various downstream benchmarks. Altogether, our findings demonstrate that CGLS unlocks the potential of progressive stacking, offering a simple yet effective strategy for improving generalization on knowledge-intensive and reasoning tasks.

cs.CL↗

arXiv2024

A deep-learning model for one-shot transcranial ultrasound simulation and phase aberration correction

Transcranial ultrasound (TUS) has emerged as a promising tool in clinical and research settings due to its potential to modulate neuronal activity, open the blood-brain barrier, facilitate targeted drug delivery via nanoparticles, and perform thermal ablation, all non-invasively. By delivering focused ultrasound waves to precise regions anywhere in the brain, TUS enables targeted energy deposition and is being explored in over fifty clinical trials as a treatment for conditions such as opioid addiction, Alzheimer's disease, dementia, epilepsy, and glioblastoma. However, effective TUS treatment requires careful ultrasound parameter design and precise computation of the focal spot's location and pressure, as skull heterogeneity increases the risk of off-target sonication or insufficient energy delivery to neural tissue. In clinical settings, this phase aberration correction must be computed within seconds. To achieve this, commercial devices often rely on faster methods, such as ray tracing, to predict the focus location and pressure. While computationally efficient, these methods may not always provide the high level of accuracy needed for optimal TUS delivery. We present TUSNet, the first end-to-end deep learning approach to solve for both the pressure field and phase aberration corrections without being bound to the inherent trade-off between accuracy and efficiency. TUSNet computes the 2D transcranial ultrasound pressure field and phase corrections within 21 milliseconds (over $1200\times$ faster than k-Wave, a MATLAB-based acoustic simulation package), achieving $98.3\%$ accuracy in estimating peak pressure magnitude at the focal spot with a mean positioning error of only $0.18$ mm compared to ground truth from k-Wave.

physics.med-ph↗

arXiv2024

Synthetic Skull CT Generation with Generative Adversarial Networks to Train Deep Learning Models for Clinical Transcranial Ultrasound

Deep learning offers potential for various healthcare applications, yet requires extensive datasets of curated medical images where data privacy, cost, and distribution mismatch across various acquisition centers could become major problems. To overcome these challenges, we propose a generative adversarial network (SkullGAN) to create large datasets of synthetic skull CT slices, geared towards training models for transcranial ultrasound. With wide ranging applications in treatment of essential tremor, Parkinson's, and Alzheimer's disease, transcranial ultrasound clinical pipelines can be significantly optimized via integration of deep learning. The main roadblock is the lack of sufficient skull CT slices for the purposes of training, which SkullGAN aims to address. Actual CT slices of 38 healthy subjects were used for training. The generated synthetic skull images were then evaluated based on skull density ratio, mean thickness, and mean intensity. Their fidelity was further analyzed using t-distributed stochastic neighbor embedding (t-SNE), Fréchet inception distance (FID) score, and visual Turing test (VTT) taken by four staff clinical radiologists. SkullGAN-generated images demonstrated similar quantitative radiological features to real skulls. t-SNE failed to separate real and synthetic samples from one another, and the FID score was 49. Expert radiologists achieved a 60\% mean accuracy on the VTT. SkullGAN makes it possible for researchers to generate large numbers of synthetic skull CT segments, necessary for training neural networks for medical applications involving the human skull, such as transcranial focused ultrasound, mitigating challenges with access, privacy, capital, time, and the need for domain expertise.

eess.IV↗

arXiv2023

New Evaluation Metrics Capture Quality Degradation due to LLM Watermarking

With the increasing use of large-language models (LLMs) like ChatGPT, watermarking has emerged as a promising approach for tracing machine-generated content. However, research on LLM watermarking often relies on simple perplexity or diversity-based measures to assess the quality of watermarked text, which can mask important limitations in watermarking. Here we introduce two new easy-to-use methods for evaluating watermarking algorithms for LLMs: 1) evaluation by LLM-judger with specific guidelines; and 2) binary classification on text embeddings to distinguish between watermarked and unwatermarked text. We apply these methods to characterize the effectiveness of current watermarking techniques. Our experiments, conducted across various datasets, reveal that current watermarking methods are detectable by even simple classifiers, challenging the notion of watermarking subtlety. We also found, through the LLM judger, that watermarking impacts text quality, especially in degrading the coherence and depth of the response. Our findings underscore the trade-off between watermark robustness and text quality and highlight the importance of having more informative metrics to assess watermarking quality.

cs.CL↗