Grokking Modular Polynomials

no code implementations • 5 Jun 2024 • Darshil Doshi, Tianyu He, Aritra Das, Andrey Gromov

Neural networks readily learn a subset of the modular arithmetic tasks, while failing to generalize on the rest.

Learning to grok: Emergence of in-context learning and skill composition in modular arithmetic tasks

1 code implementation • 4 Jun 2024 • Tianyu He, Darshil Doshi, Aritra Das, Andrey Gromov

In this work, we study the emergence of in-context learning and skill composition in a collection of modular arithmetic tasks.

In-Context Learning Out-of-Distribution Generalization

To grok or not to grok: Disentangling generalization and memorization on corrupted algorithmic datasets

1 code implementation • 19 Oct 2023 • Darshil Doshi, Aritra Das, Tianyu He, Andrey Gromov

Robust generalization is a major challenge in deep learning, particularly when the number of trainable parameters is very large.

Memorization

An India-specific Compartmental Model for Covid-19: Projections and Intervention Strategies by Incorporating Geographical, Infrastructural and Response Heterogeneity

2 code implementations • 28 Jul 2020 • Sanit Gupta, Sahil Shah, Sumit Chaturvedi, Pranav Thakkar, Parvinder Solanki, Soham Dibyachintan, Sandeepan Roy, M. B. Sushma, Adwait Godbole, Noufal Jaseem, Pradumn Kumar, Sucheta Ravikanti, Aritra Das, Giridhara R. Babu, Tarun Bhatnagar, Avijit Maji, Mithun K. Mitra, Sai Vinjanampathy

We present a compartmental meta-population model for the spread of Covid-19 in India.

Combining Multi-level Contexts of Superpixel using Convolutional Neural Networks to perform Natural Scene Labeling

no code implementations • 14 Mar 2018 • Aritra Das, Swarnendu Ghosh, Ritesh Sarkhel, Sandipan Choudhuri, Nibaran Das, Mita Nasipuri

Modern deep learning algorithms have triggered various image segmentation approaches.

Image Segmentation Scene Labeling +3