Alert button
Picture for Rishabh Iyer

Rishabh Iyer

Alert button

STENCIL: Submodular Mutual Information Based Weak Supervision for Cold-Start Active Learning

Feb 21, 2024
Nathan Beck, Adithya Iyer, Rishabh Iyer

Viaarxiv icon

Theoretical Analysis of Submodular Information Measures for Targeted Data Subset Selection

Feb 21, 2024
Nathan Beck, Truong Pham, Rishabh Iyer

Viaarxiv icon

Gradient Coreset for Federated Learning

Jan 13, 2024
Durga Sivasubramanian, Lokesh Nagalapatti, Rishabh Iyer, Ganesh Ramakrishnan

Viaarxiv icon

SCoRe: Submodular Combinatorial Representation Learning for Real-World Class-Imbalanced Settings

Sep 29, 2023
Anay Majee, Suraj Kothawade, Krishnateja Killiamsetty, Rishabh Iyer

Figure 1 for SCoRe: Submodular Combinatorial Representation Learning for Real-World Class-Imbalanced Settings
Figure 2 for SCoRe: Submodular Combinatorial Representation Learning for Real-World Class-Imbalanced Settings
Figure 3 for SCoRe: Submodular Combinatorial Representation Learning for Real-World Class-Imbalanced Settings
Figure 4 for SCoRe: Submodular Combinatorial Representation Learning for Real-World Class-Imbalanced Settings
Viaarxiv icon

Beyond Active Learning: Leveraging the Full Potential of Human Interaction via Auto-Labeling, Human Correction, and Human Verification

Jun 02, 2023
Nathan Beck, Krishnateja Killamsetty, Suraj Kothawade, Rishabh Iyer

Figure 1 for Beyond Active Learning: Leveraging the Full Potential of Human Interaction via Auto-Labeling, Human Correction, and Human Verification
Figure 2 for Beyond Active Learning: Leveraging the Full Potential of Human Interaction via Auto-Labeling, Human Correction, and Human Verification
Figure 3 for Beyond Active Learning: Leveraging the Full Potential of Human Interaction via Auto-Labeling, Human Correction, and Human Verification
Figure 4 for Beyond Active Learning: Leveraging the Full Potential of Human Interaction via Auto-Labeling, Human Correction, and Human Verification
Viaarxiv icon

STREAMLINE: Streaming Active Learning for Realistic Multi-Distributional Settings

May 18, 2023
Nathan Beck, Suraj Kothawade, Pradeep Shenoy, Rishabh Iyer

Figure 1 for STREAMLINE: Streaming Active Learning for Realistic Multi-Distributional Settings
Figure 2 for STREAMLINE: Streaming Active Learning for Realistic Multi-Distributional Settings
Figure 3 for STREAMLINE: Streaming Active Learning for Realistic Multi-Distributional Settings
Figure 4 for STREAMLINE: Streaming Active Learning for Realistic Multi-Distributional Settings
Viaarxiv icon

INGENIOUS: Using Informative Data Subsets for Efficient Pre-Training of Large Language Models

May 11, 2023
H S V N S Kowndinya Renduchintala, Krishnateja Killamsetty, Sumit Bhatia, Milan Aggarwal, Ganesh Ramakrishnan, Rishabh Iyer, Balaji Krishnamurthy

Figure 1 for INGENIOUS: Using Informative Data Subsets for Efficient Pre-Training of Large Language Models
Figure 2 for INGENIOUS: Using Informative Data Subsets for Efficient Pre-Training of Large Language Models
Figure 3 for INGENIOUS: Using Informative Data Subsets for Efficient Pre-Training of Large Language Models
Figure 4 for INGENIOUS: Using Informative Data Subsets for Efficient Pre-Training of Large Language Models
Viaarxiv icon

MILO: Model-Agnostic Subset Selection Framework for Efficient Model Training and Tuning

Feb 05, 2023
Krishnateja Killamsetty, Alexandre V. Evfimievski, Tejaswini Pedapati, Kiran Kate, Lucian Popa, Rishabh Iyer

Figure 1 for MILO: Model-Agnostic Subset Selection Framework for Efficient Model Training and Tuning
Figure 2 for MILO: Model-Agnostic Subset Selection Framework for Efficient Model Training and Tuning
Figure 3 for MILO: Model-Agnostic Subset Selection Framework for Efficient Model Training and Tuning
Figure 4 for MILO: Model-Agnostic Subset Selection Framework for Efficient Model Training and Tuning
Viaarxiv icon

Partitioned Gradient Matching-based Data Subset Selection for Compute-Efficient Robust ASR Training

Oct 30, 2022
Ashish Mittal, Durga Sivasubramanian, Rishabh Iyer, Preethi Jyothi, Ganesh Ramakrishnan

Figure 1 for Partitioned Gradient Matching-based Data Subset Selection for Compute-Efficient Robust ASR Training
Figure 2 for Partitioned Gradient Matching-based Data Subset Selection for Compute-Efficient Robust ASR Training
Figure 3 for Partitioned Gradient Matching-based Data Subset Selection for Compute-Efficient Robust ASR Training
Figure 4 for Partitioned Gradient Matching-based Data Subset Selection for Compute-Efficient Robust ASR Training
Viaarxiv icon