A5103120216- Artificial Intelligence in Healthcare
- Topic Modeling
- Anomaly Detection Techniques and Applications
- Ethics and Social Impacts of AI
- Text Readability and Simplification
- ECG Monitoring and Analysis
- Time Series Analysis and Forecasting
- Multimodal Machine Learning Applications
- Electronic Health Records Systems
- Natural Language Processing Techniques
- Technology and Data Analysis
- Brain Tumor Detection and Classification
- Complex Network Analysis Techniques
- Innovation in Digital Healthcare Systems
- Building Energy and Comfort Optimization
- Advanced Manufacturing and Logistics Optimization
- Adversarial Robustness in Machine Learning
- Pharmacy and Medical Practices
- Engineering Applied Research
- Supply Chain and Inventory Management
- Cloud Computing and Resource Management
- Advanced Bandit Algorithms Research
- Emotion and Mood Recognition
- AI in Service Interactions
- Digital Transformation in Industry
Parul University
2024-2025
Amazon (United States)
2022-2023
Harvard University Press
2022
Georgia Institute of Technology
2016-2017
In this work, we demonstrate that multilingual large-scale sequence-to-sequence (seq2seq) models, pre-trained on a mixture of denoising and Causal Language Modeling (CLM) tasks, are more efficient few-shot learners than decoder-only models various tasks. particular, train 20 billion parameter seq2seq model called Alexa Teacher Model (AlexaTM 20B) show it achieves state-of-the-art (SOTA) performance 1-shot summarization outperforming much larger 540B PaLM decoder model. AlexaTM 20B also SOTA...
Large volumes of networked streaming event data are becoming increasingly available in a wide variety applications such as social network analysis, Internet traffic monitoring, and health care analytics. Streaming discrete observations occurring continuous time, the precise time interval between two events carries substantial information about dynamics underlying systems. How does one promptly detect changes these dynamic systems using data? In this paper, we propose novel change-point...
Umang Gupta, Jwala Dhamala, Varun Kumar, Apurv Verma, Yada Pruksachatkun, Satyapriya Krishna, Rahul Kai-Wei Chang, Greg Ver Steeg, Aram Galstyan. Findings of the Association for Computational Linguistics: ACL 2022.
Creating secure and resilient applications with large language models (LLM) requires anticipating, adjusting to, countering unforeseen threats. Red-teaming has emerged as a critical technique for identifying vulnerabilities in real-world LLM implementations. This paper presents detailed threat model provides systematization of knowledge (SoK) red-teaming attacks on LLMs. We develop taxonomy based the stages development deployment process extract various insights from previous research. In...
Large volume of networked streaming event data are becoming increasingly available in a wide variety applications, such as social network analysis, Internet traffic monitoring and healthcare analytics. Streaming discrete observation occurred continuous time, the precise time interval between two events carries great deal information about dynamics underlying systems. How to promptly detect changes these dynamic systems using data? In this paper, we propose novel change-point detection...
Recently, there is a surge of interest in using point processes to model continuous-time user activities. This framework has resulted novel models and improved performance diverse applications. However, most previous works focus on the "open loop" setting where learned are used for predictive tasks. Typically, we interested "closed policy needs be incorporate feedbacks guide activities desirable states. Although have good performance, it not clear how use them challenging closed loop...
Ninareh Mehrabi, Palash Goyal, Apurv Verma, Jwala Dhamala, Varun Kumar, Qian Hu, Kai-Wei Chang, Richard Zemel, Aram Galstyan, Rahul Gupta. Proceedings of the 61st Annual Meeting Association for Computational Linguistics (Volume 1: Long Papers). 2023.
Natural language often contains ambiguities that can lead to misinterpretation and miscommunication. While humans handle effectively by asking clarifying questions and/or relying on contextual cues common-sense knowledge, resolving be notoriously hard for machines. In this work, we study arise in text-to-image generative models. We curate a benchmark dataset covering different types of occur these systems. then propose framework mitigate the prompts given systems soliciting clarifications...
Satyapriya Krishna, Rahul Gupta, Apurv Verma, Jwala Dhamala, Yada Pruksachatkun, Kai-Wei Chang. Proceedings of the 60th Annual Meeting Association for Computational Linguistics (Volume 1: Long Papers). 2022.