arXiv subjects

Shashank Yadav

Publications and source records attributed to Shashank Yadav.

13 recordsLinked to original sources

arXiv2026

Morphology signal in whole slide image foundation models can automatically triage slides

Patient exams in the cancer diagnosis and staging process typically generate several whole slide images (WSIs). One of the initial steps in training models on WSI data is identifying one or a few slides containing tumor or other diagnostic biomarkers necessary for downstream prediction tasks such as estimating recurrence risk or progression-free survival. This step requires tedious manual curation by experienced pathologists. Many published datasets make the artificial assumption of 1 slide per patient. Alternatively, all slides per patient may be used for model training, which may dilute the signal from the few slides containing tumor or other relevant information. In this paper, we present a pipeline to overcome these challenges using publicly available WSI foundation models (FMs). Our evaluations show that ranking WSIs based on predictions from zero-shot classification using WSI FMs accurately identifies slides with the most tumor, indicating that WSI FMs contain sufficient morphology signal to automatically triage slides. We also present a formulation for ranked evaluation to benchmark FM performance in slide triage. We show, on multiple datasets, that tumor slides are identified in the top-2 ranked slides for patients with up to 43 slides.

cs.CV↗

arXiv2026

The Cost of Binarizing Survival Outcomes in Clinical Prognostic Modeling

Survival analysis is an established framework for analyzing time-to-event data, yet many clinical machine learning studies still binarize the outcome before model training. This practice excludes censored patients, collapses temporal information into a single threshold, and can affect which features are selected as prognostically relevant. We examine the cost of this binarization in the context of Bayesian network (BN) feature selection, using two recent publications as case studies: one that applies BN-based feature selection to a head-and-neck cancer cohort and a second surgical cohort study that, while not BN-based, likewise binarizes its survival endpoint. We replace the binary scoring function with the Cox partial log-likelihood for feature-to-outcome edges, a modification we call the Survival-Aware Bayesian network, and recover prognostic features that binarization misses. Our ablation experiment confirms that the improvement is driven by the time-to-event scoring formulation rather than by retaining more patients. The results generalize across five endpoint-cohort combinations in head-and-neck cancer and extend to three further cancer types (breast, colorectal, and kidney). We propose that clinical studies with survival outcomes should use time-to-event methods by default, as binarization discards the prognostic signal retained by survival analysis.

q-bio.QM↗

arXiv2025

Reflections from Research Roundtables at the Conference on Health, Inference, and Learning (CHIL) 2025

The 6th Annual Conference on Health, Inference, and Learning (CHIL 2025), hosted by the Association for Health Learning and Inference (AHLI), was held in person on June 25-27, 2025, at the University of California, Berkeley, in Berkeley, California, USA. As part of this year's program, we hosted Research Roundtables to catalyze collaborative, small-group dialogue around critical, timely topics at the intersection of machine learning and healthcare. Each roundtable was moderated by a team of senior and junior chairs who fostered open exchange, intellectual curiosity, and inclusive engagement. The sessions emphasized rigorous discussion of key challenges, exploration of emerging opportunities, and collective ideation toward actionable directions in the field. In total, eight roundtables were held by 19 roundtable chairs on topics of "Explainability, Interpretability, and Transparency," "Uncertainty, Bias, and Fairness," "Causality," "Domain Adaptation," "Foundation Models," "Learning from Small Medical Data," "Multimodal Methods," and "Scalable, Translational Healthcare Solutions."

cs.LG↗

arXiv2025

Lightweight Language Models are Prone to Reasoning Errors for Complex Computational Phenotyping Tasks

Although computational phenotyping is a central informatics activity with resulting cohorts supporting a wide variety of applications, it is time-intensive because of manual data review. We previously assessed the ability of LLMs to perform computational phenotyping tasks using computable phenotypes for ARF respiratory support therapies. They successfully performed concept classification and classification of single-therapy phenotypes but underperformed on multi-therapy phenotypes. To better understand issues with these complex tasks, we expanded PHEONA, a generalizable framework for evaluation of LLMs, to include methods specifically for evaluating faulty reasoning. We assessed the responses of two lightweight non-reasoning LLMs (Mistral Small 24 billion and Phi-4 14 billion) and one lightweight reasoning LLM (Qwen-distilled DeepSeek-r1 32 billion) both with and without prompt modifications to identify explanation correctness errors and unfaithfulness errors during phenotyping. For experiments without prompt modifications, both errors were present in responses from all models. For experiments with prompt modifications, we measured the mean absolute change in accuracy relative to the unbiased prompt across biasing conditions. Adding specific few-shot examples aligned with an incorrect phenotype reduced accuracy by at least 5% and up to 10% depending on the model and CoT type. Since reasoning errors were ubiquitous across models, our enhancement of PHEONA to include a component for assessing faulty reasoning provides a practical framework for evaluating LLM reasoning and empirical evidence that reasoning errors occur during complex computational phenotyping.

q-bio.QM↗

arXiv2025

Failure Modes of Time Series Interpretability Algorithms for Critical Care Applications and Potential Solutions

Interpretability plays a vital role in aligning and deploying deep learning models in critical care, especially in constantly evolving conditions that influence patient survival. However, common interpretability algorithms face unique challenges when applied to dynamic prediction tasks, where patient trajectories evolve over time. Gradient, Occlusion, and Permutation-based methods often struggle with time-varying target dependency and temporal smoothness. This work systematically analyzes these failure modes and supports learnable mask-based interpretability frameworks as alternatives, which can incorporate temporal continuity and label consistency constraints to learn feature importance over time. Here, we propose that learnable mask-based approaches for dynamic timeseries prediction problems provide more reliable and consistent interpretations for applications in critical care and similar domains.

cs.LG↗

arXiv2025

SHREC: A Framework for Advancing Next-Generation Computational Phenotyping with Large Language Models

Computational phenotyping is a central informatics activity with resulting cohorts supporting a wide variety of applications. However, it is time-intensive because of manual data review and limited automation. Since LLMs have demonstrated promising capabilities for text classification, comprehension, and generation, we posit they will perform well at repetitive manual review tasks traditionally performed by human experts. To support next-generation computational phenotyping, we developed SHREC, a framework for integrating LLMs into end-to-end phenotyping pipelines. We applied and tested three lightweight LLMs (Gemma2 27 billion, Mistral Small 24 billion, and Phi-4 14 billion) to classify concepts and phenotype patients using phenotypes for ARF respiratory support therapies. All models performed well on concept classification, with the best (Mistral) achieving an AUROC of 0.896. For phenotyping, models demonstrated near-perfect specificity for all phenotypes with the top-performing model (Mistral) achieving an average AUROC of 0.853 for single-therapy phenotypes. In conclusion, lightweight LLMs can assist researchers with resource-intensive phenotyping tasks. Several advantages of LLMs included their ability to adapt to new tasks with prompt engineering alone and their ability to incorporate raw EHR data. Future steps include determining optimal strategies for integrating biomedical data and understanding reasoning errors

q-bio.QM↗

arXiv2025

PHEONA: An Evaluation Framework for Large Language Model-based Approaches to Computational Phenotyping

Computational phenotyping is essential for biomedical research but often requires significant time and resources, especially since traditional methods typically involve extensive manual data review. While machine learning and natural language processing advancements have helped, further improvements are needed. Few studies have explored using Large Language Models (LLMs) for these tasks despite known advantages of LLMs for text-based tasks. To facilitate further research in this area, we developed an evaluation framework, Evaluation of PHEnotyping for Observational Health Data (PHEONA), that outlines context-specific considerations. We applied and demonstrated PHEONA on concept classification, a specific task within a broader phenotyping process for Acute Respiratory Failure (ARF) respiratory support therapies. From the sample concepts tested, we achieved high classification accuracy, suggesting the potential for LLM-based methods to improve computational phenotyping processes.

cs.CL↗

arXiv2024

Gamified crowd-sourcing of high-quality data for visual fine-tuning

This paper introduces Gamified Adversarial Prompting (GAP), a framework that crowd-sources high-quality data for visual instruction tuning of large multimodal models. GAP transforms the data collection process into an engaging game, incentivizing players to provide fine-grained, challenging questions and answers that target gaps in the model's knowledge. Our contributions include (1) an approach to capture question-answer pairs from humans that directly address weaknesses in a model's knowledge, (2) a method for evaluating and rewarding players that successfully incentivizes them to provide high-quality submissions, and (3) a scalable, gamified platform that succeeds in collecting this data from over 50,000 participants in just a few weeks. Our implementation of GAP has significantly improved the accuracy of a small multimodal model, namely MiniCPM-Llama3-V-2.5-8B, increasing its GPT score from 0.147 to 0.477 on our dataset, approaching the benchmark set by the much larger GPT-4V. Moreover, we demonstrate that the data generated using MiniCPM-Llama3-V-2.5-8B also enhances its performance across other benchmarks, and exhibits cross-model benefits. Specifically, the same data improves the performance of QWEN2-VL-2B and QWEN2-VL-7B on the same multiple benchmarks.

cs.AI↗

arXiv2023

Socialbots and the Challenges of Cyberspace Awareness

As security communities brace for the emerging social automation based threats, we examine the mechanisms of developing situation awareness in cyberspace and the governance issues that socialbots bring into this existing paradigm of cyber situation awareness. We point out that an organisation's situation awareness in cyberspace is a phenomena fundamentally distinct from the original conception of situation awareness, requiring continuous data exchange and knowledge management where the standard implementation mechanisms require significant policy attention in light of threats like malicious social automation. We conceptualise Cyberspace Awareness as a socio-technical phenomena with Syntactic, Semantic, and Operatic dimensions - each subject to a number of stressors which are exacerbated under social automation based threats. The paper contributes to the ideas of situational awareness in cyberspace, and characterises the challenges therein around tackling the increasingly social and often pervasive, automation in cyber threat environments.

cs.CY↗

arXiv2022

Political Propagation of Social Botnets: Policy Consequences

The 2016 US election was a watershed event where an electoral intervention by an adversarial state made extensive use of networks of software robots and data driven communications which transformed the interference into a goal driven functionality of man-machine collaboration. Reviewing the debates post the debacle, we reflect upon the policy consequences of the use of Social Botnets and understand the impact of their adversarial operation in terms of catalysing institutional decay, growing infrastructural anxieties, increased industry regulations, more vulnerable Individuals and more distorted ideas, and most importantly, the emergence of an unintended constituency in form of the bot agency itself. The article first briefly introduces the nature and evolution of Social Botnets, and then moves over to discussing the policy consequences. For future work, it is important to understand the agency and collective properties of these software robots, in order to design the institutional and socio-technical mechanisms which mitigate the risk of adversarial social engineering using these bots from interfering into democratic processes.

cs.CY↗

arXiv2022

Machines & Influence: An Information Systems Lens

Policymakers face a broader challenge of how to view AI capabilities today and where does society stand in terms of those capabilities. This paper surveys AI capabilities and tackles this very issue, exploring it in context of political security in digitally networked societies. We extend the ideas of Information Management to better understand contemporary AI systems as part of a larger and more complex information system. Comprehensively reviewing AI capabilities and contemporary man-machine interactions, we undertake conceptual development to suggest that better information management could allow states to more optimally offset the risks of AI enabled influence and better utilise the emerging capabilities which these systems have to offer to policymakers and political institutions across the world. Hopefully this long essay will actuate further debates and discussions over these ideas, and prove to be a useful contribution towards governing the future of AI.

cs.CY↗

arXiv2019

Structuring an unordered text document

Segmenting an unordered text document into different sections is a very useful task in many text processing applications like multiple document summarization, question answering, etc. This paper proposes structuring of an unordered text document based on the keywords in the document. We test our approach on Wikipedia documents using both statistical and predictive methods such as the TextRank algorithm and Google's USE (Universal Sentence Encoder). From our experimental results, we show that the proposed model can effectively structure an unordered document into sections.

cs.IR↗

arXiv2018

A Joint 3D-2D based Method for Free Space Detection on Roads

In this paper, we address the problem of road segmentation and free space detection in the context of autonomous driving. Traditional methods either use 3-dimensional (3D) cues such as point clouds obtained from LIDAR, RADAR or stereo cameras or 2-dimensional (2D) cues such as lane markings, road boundaries and object detection. Typical 3D point clouds do not have enough resolution to detect fine differences in heights such as between road and pavement. Image based 2D cues fail when encountering uneven road textures such as due to shadows, potholes, lane markings or road restoration. We propose a novel free road space detection technique combining both 2D and 3D cues. In particular, we use CNN based road segmentation from 2D images and plane/box fitting on sparse depth data obtained from SLAM as priors to formulate an energy minimization using conditional random field (CRF), for road pixels classification. While the CNN learns the road texture and is unaffected by depth boundaries, the 3D information helps in overcoming texture based classification failures. Finally, we use the obtained road segmentation with the 3D depth data from monocular SLAM to detect the free space for the navigation purposes. Our experiments on KITTI odometry dataset, Camvid dataset, as well as videos captured by us, validate the superiority of the proposed approach over the state of the art.

cs.CV↗