Sirisha Rambhatla
YOU?
Author Swipe
View article: Randomized Gradient Subspaces for Efficient Large Language Model Training
Randomized Gradient Subspaces for Efficient Large Language Model Training Open
Training large language models (LLMs) is often bottlenecked by extreme memory demands, with optimizer states dominating the footprint. Recent works mitigates this cost by projecting gradients into low-dimensional subspaces using sophistica…
View article: Bringing Equity to Classification: Domain Generalization for Domain-Linked Classes
Bringing Equity to Classification: Domain Generalization for Domain-Linked Classes Open
Domain generalization (DG) focuses on transferring domain-invariant knowledge from multiple source (training) domains to an a priori unseen target domain(s). This task implicitly requires that classes of interest are expressed in multiple …
View article: Leveraging large language models for automated depression screening.
Leveraging large language models for automated depression screening. Open
Mental health diagnoses possess unique challenges that often lead to nuanced difficulties in managing an individual's well-being and daily functioning. Self-report questionnaires are a common practice in clinical settings to help mitigate …
View article: SafeTuneBed: A Toolkit for Benchmarking LLM Safety Alignment in Fine-Tuning
SafeTuneBed: A Toolkit for Benchmarking LLM Safety Alignment in Fine-Tuning Open
As large language models (LLMs) become ubiquitous, parameter-efficient fine-tuning methods and safety-first defenses have proliferated rapidly. However, the number of approaches and their recent increase have resulted in diverse evaluation…
View article: Understanding LLM Scientific Reasoning through Promptings and Model's Explanation on the Answers
Understanding LLM Scientific Reasoning through Promptings and Model's Explanation on the Answers Open
Large language models (LLMs) have demonstrated remarkable capabilities in natural language understanding, reasoning, and problem-solving across various domains. However, their ability to perform complex, multi-step reasoning task-essential…
View article: Embedding-Based Representation Learning for Forecasting Flight Characteristics
Embedding-Based Representation Learning for Forecasting Flight Characteristics Open
Airlines face significant challenges when building flight schedules, particularly because of unpredictable operations caused by factors such as adverse weather, airport congestion, mechanical problems, and so forth. One of the major compon…
View article: Opinion: Mental health research: to augment or not to augment
Opinion: Mental health research: to augment or not to augment Open
View article: SubTrack++ : Gradient Subspace Tracking for Scalable LLM Training
SubTrack++ : Gradient Subspace Tracking for Scalable LLM Training Open
Training large language models (LLMs) is highly resource-intensive due to their massive number of parameters and the overhead of optimizer states. While recent work has aimed to reduce memory consumption, such efforts often entail trade-of…
View article: Medical Misinformation in AI-Assisted Self-Diagnosis: Development of a Method (EvalPrompt) for Analyzing Large Language Models
Medical Misinformation in AI-Assisted Self-Diagnosis: Development of a Method (EvalPrompt) for Analyzing Large Language Models Open
Background Rapid integration of large language models (LLMs) in health care is sparking global discussion about their potential to revolutionize health care quality and accessibility. At a time when improving health care quality and access…
View article: Boundary-Aware Semantic Segmentation for Ice Hockey Rink Registration
Boundary-Aware Semantic Segmentation for Ice Hockey Rink Registration Open
View article: Domain-guided Masked Autoencoders for Unique Player Identification
Domain-guided Masked Autoencoders for Unique Player Identification Open
Unique player identification is a fundamental module in vision-driven sports analytics. Identifying players from broadcast videos can aid with various downstream tasks such as player assessment, in-game analysis, and broadcast production. …
View article: Domain-Guided Masked Autoencoders for Unique Player Identification
Domain-Guided Masked Autoencoders for Unique Player Identification Open
Unique player identification is a fundamental module in vision-driven sports analytics. Identifying players from broadcast videos can aid with various downstream tasks such as player assessment, in-game analysis, and broadcast production. …
View article: Domain-Guided Spatio-Temporal Self-Attention for Egocentric 3D Pose Estimation
Domain-Guided Spatio-Temporal Self-Attention for Egocentric 3D Pose Estimation Open
Vision-based ego-centric 3D human pose estimation (ego-HPE) is essential to support critical applications of xR-technologies. However, severe self-occlusions and strong distortion introduced by the fish-eye view from the head mounted camer…
View article: Medical Misinformation in AI-Assisted Self-Diagnosis: Development of a Method (EvalPrompt) for Analyzing Large Language Models
Medical Misinformation in AI-Assisted Self-Diagnosis: Development of a Method (EvalPrompt) for Analyzing Large Language Models Open
Rapid integration of large language models (LLMs) in health care is sparking global discussion about their potential to revolutionize health care quality and accessibility. At a time when improving health care quality and access remains a …
View article: Is Generative Modeling-based Stylization Necessary for Domain Adaptation in Regression Tasks?
Is Generative Modeling-based Stylization Necessary for Domain Adaptation in Regression Tasks? Open
Unsupervised domain adaptation (UDA) aims to bridge the gap between source and target domains in the absence of target domain labels using two main techniques: input-level alignment (such as generative modeling and stylization) and feature…
View article: Domain Generalization for Domain-Linked Classes
Domain Generalization for Domain-Linked Classes Open
Domain generalization (DG) focuses on transferring domain-invariant knowledge from multiple source domains (available at train time) to an, a priori, unseen target domain(s). This requires a class to be expressed in multiple domains for th…
View article: I-SEA: Importance Sampling and Expected Alignment-Based Deep Distance Metric Learning for Time Series Analysis and Embedding
I-SEA: Importance Sampling and Expected Alignment-Based Deep Distance Metric Learning for Time Series Analysis and Embedding Open
Learning effective embeddings for potentially irregularly sampled time-series, evolving at different time scales, is fundamental for machine learning tasks such as classification and clustering. Task-dependent embeddings rely on similariti…
View article: Building Spatio-temporal Transformers for Egocentric 3D Pose Estimation
Building Spatio-temporal Transformers for Egocentric 3D Pose Estimation Open
Egocentric 3D human pose estimation (HPE) from images is challenging due to severe self-occlusions and strong distortion introduced by the fish-eye view from the head mounted camera. Although existing works use intermediate heatmap-based r…
View article: Toward Accurate Spatiotemporal COVID-19 Risk Scores Using High-Resolution Real-World Mobility Data
Toward Accurate Spatiotemporal COVID-19 Risk Scores Using High-Resolution Real-World Mobility Data Open
As countries look toward re-opening of economic activities amidst the ongoing COVID-19 pandemic, ensuring public health has been challenging. While contact tracing only aims to track past activities of infected users, one path to safe reop…
View article: Cross-Node Federated Graph Neural Network for Spatio-Temporal Data Modeling
Cross-Node Federated Graph Neural Network for Spatio-Temporal Data Modeling Open
Vast amount of data generated from networks of sensors, wearables, and the Internet of Things (IoT) devices underscores the need for advanced modeling techniques that leverage the spatio-temporal structure of decentralized data due to the …
View article: PD58-08 AUTOMATING SUTURING SKILLS ASSESSMENT WITH A LIMITED SURGEON DATASET: META LEARNING
PD58-08 AUTOMATING SUTURING SKILLS ASSESSMENT WITH A LIMITED SURGEON DATASET: META LEARNING Open
You have accessJournal of UrologySurgical Technology & Simulation: Training & Skills Assessment (PD58)1 Sep 2021PD58-08 AUTOMATING SUTURING SKILLS ASSESSMENT WITH A LIMITED SURGEON DATASET: META LEARNING Andrew J. Hung, Sirisha Rambhatla, …
View article: Physics-aware Spatiotemporal Modules with Auxiliary Tasks for Meta-Learning
Physics-aware Spatiotemporal Modules with Auxiliary Tasks for Meta-Learning Open
Modeling the dynamics of real-world physical systems is critical for spatiotemporal prediction tasks, but challenging when data is limited. The scarcity of real-world data and the difficulty in reproducing the data distribution hinder dire…
View article: PolSIRD: Modeling Epidemic Spread Under Intervention Policies
PolSIRD: Modeling Epidemic Spread Under Intervention Policies Open
View article: Cross-Node Federated Graph Neural Network for Spatio-Temporal Data Modeling
Cross-Node Federated Graph Neural Network for Spatio-Temporal Data Modeling Open
Vast amount of data generated from networks of sensors, wearables, and the Internet of Things (IoT) devices underscores the need for advanced modeling techniques that leverage the spatio-temporal structure of decentralized data due to the …
View article: Cross-Node Federated Graph Neural Network for Spatio-Temporal Data Modeling
Cross-Node Federated Graph Neural Network for Spatio-Temporal Data Modeling Open
Dataset and code for "Cross-Node Federated Graph Neural Network for Spatio-Temporal Data Modeling".
View article: Cross-Node Federated Graph Neural Network for Spatio-Temporal Data Modeling
Cross-Node Federated Graph Neural Network for Spatio-Temporal Data Modeling Open
Dataset and code for "Cross-Node Federated Graph Neural Network for Spatio-Temporal Data Modeling".
View article: DL4Burn: Burn Surgical Candidacy Prediction using Multimodal Deep Learning.
DL4Burn: Burn Surgical Candidacy Prediction using Multimodal Deep Learning. Open
Burn wounds are most commonly evaluated through visual inspection to determine surgical candidacy, taking into account burn depth and individualized patient factors. This process, though cost effective, is subjective and varies by provider…
View article: Towards Accurate Spatiotemporal COVID-19 Risk Scores using High Resolution Real-World Mobility Data
Towards Accurate Spatiotemporal COVID-19 Risk Scores using High Resolution Real-World Mobility Data Open
As countries look towards re-opening of economic activities amidst the ongoing COVID-19 pandemic, ensuring public health has been challenging. While contact tracing only aims to track past activities of infected users, one path to safe reo…
View article: PolSIRD: Modeling Epidemic Spread under Intervention Policies and an Application to the Spread of COVID-19
PolSIRD: Modeling Epidemic Spread under Intervention Policies and an Application to the Spread of COVID-19 Open
Epidemic spread in a population is traditionally modeled via compartmentalized models which represent the free evolution of disease in absence of any intervention policies. In addition, these models assume full observability and do not acc…
View article: Provable Online CP/PARAFAC Decomposition of a Structured Tensor via Dictionary Learning
Provable Online CP/PARAFAC Decomposition of a Structured Tensor via Dictionary Learning Open
We consider the problem of factorizing a structured 3-way tensor into its constituent Canonical Polyadic (CP) factors. This decomposition, which can be viewed as a generalization of singular value decomposition (SVD) for tensors, reveals h…