Hi! I am Amrith, a graduate student at the School of Computer Science, Carnegie Mellon University pursuing a Masters degree in Language Technologies. I am fortunate to be advised by professors from both MLD and LTI: Prof. Virginia Smith (MLD), Prof. Barnabas Poczos (MLD) and Prof. Alan W. Black (LTI). My primary research interest lies in bridging the gap between machine learning theory and practice. In search of this broad goal I am currently trying to theoretically understand the empirical successes of deep learning in NLP/Vision while also striving to make these models more interpretable, computationally efficient and amenable for fast adaptation. I am particularly interested in statistical ML, representation learning, large-scale optimization, multi-task/meta-learning, Bayesian methods, generative models in text/speech, causal inference for text.


asetlur [at] cs.cmu[dot]edu;

asetlur [at] andrew[dot]


Lessons from Chasing Few-Shot Learning Benchmarks:
Rethinking the Evaluation of Meta-Learning Methods

Amrith Setlur*, Oscar Li*, Virginia Smith.

Explaining The Efficacy of Counterfactually-Augmented Data

Divyansh Kaushik, Amrith Setlur, Eduard Hovy, Zachary C. Lipton.
To be presented at Inernational Conference on Learning Representations (ICLR 2021)

Is Support Set Diversity Necessary for Meta-learning?

Amrith Setlur*, Oscar Li*, Virginia Smith.
Workshop on Meta-learning NeurIPS 2020

Nonlinear ISA with Auxiliary Variables for Learning Speech Representations

(Finalist for the Best Student Paper Award)

Amrith Setlur, Barnabas Poczos, Alan W Black.
In the proceedings of Interspeech 2020

Politeness Transfer: A Tag and Generate Approach

Aman Madaan*, Amrith Setlur*, Tanmay Parekh*, Barnabas Poczos, Graham Neubig, Yiming Yang,
Ruslan Salakhutdinov, Alan W Black, Shrimai Prabhumoye.
In the proceedings of Association for Computational Linguistics Conference (ACL 2020) (Long Paper).

Featured in multiple news outlets: SCS CMU News, TechCrunch, CNET, Pittsburgh Post-Gazette, msn, Hindustan Times, and Axios.

Covariate Distribution Aware Meta-learning

Amrith Setlur*, Saket Dingliwal*, Barnabas Poczos.
Workshop on Lifelong Learning ICML 2020

Better Approximate Inference for Partial Likelihood Models with a Latent Structure

(Selected for Oral Presentation)

Amrith Setlur, Barnabas Poczos.
Workshop on Temporal Point Processes NeurIPS 2019

ReStGAN: A step towards visually guided shopper experience via text-to-image synthesis

S.Surya, Amrith Setlur, Arijit Biswas, Sumit Negi
IEEE & CVF Winter Conference on Applications in Computer Vision (WACV 2020)

An efficient fault tolerant workflow scheduling approach using replication heuristics and checkpointing in the cloud

Amrith Setlur, S. Nirmala, H. Singh, S. Khoriya.
Journal of Parallel and Distributed Computing (JPDC) (Vol. 136 Feb 2020)

Robust Handwriting Recognition with Limited and Noisy Data

H. Pham Amrith Setlur, Saket Dingliwal, Tzu-Hsiang Lin, Barnabas Poczos, Kang Huang, Zhuo Li, Jae Lim
International Conference on Frontiers of Handwriting Recognition (ICFHR 2020)

Copyright © Amrith Setlur