Vishaal Udandarao's picture

Vishaal Udandarao

vishaal27

·

https://vishaal27.github.io/

AI & ML interests

Data-centric perspectives on generalisation, foundation models, multimodal learning

Recent Activity

liked a model 15 days ago

mlfoundations/fasttext-oh-eli5

liked a dataset 17 days ago

EssentialAI/essential-web-v1.0

upvoted a collection 25 days ago

DatologyAI CLIP Models

View all activity

Organizations

authored a paper 3 months ago

A Sober Look at Progress in Language Model Reasoning: Pitfalls and Paths to Reproducibility

Paper • 2504.07086 • Published Apr 9 • 21

authored a paper 7 months ago

ONEBench to Test Them All: Sample-Level Benchmarking Over Open-Ended Capabilities

Paper • 2412.06745 • Published Dec 9, 2024 • 6

authored a paper 11 months ago

Data Contamination Report from the 2024 CONDA Shared Task

Paper • 2407.21530 • Published Jul 31, 2024 • 10

authored 4 papers about 1 year ago

No "Zero-Shot" Without Exponential Data: Pretraining Concept Frequency Determines Multimodal Model Performance

Paper • 2404.04125 • Published Apr 4, 2024 • 30

Lifelong Benchmarks: Efficient Model Evaluation in an Era of Rapid Progress

Paper • 2402.19472 • Published Feb 29, 2024 • 2

Visual Data-Type Understanding does not emerge from Scaling Vision-Language Models

Paper • 2310.08577 • Published Oct 12, 2023 • 1

SuS-X: Training-Free Name-Only Transfer of Vision-Language Models

Paper • 2211.16198 • Published Nov 28, 2022 • 1