YerevaNN

research lab

YerevaNN /jɛɾɛvɑnˈɛn/ is a machine learning research center based in Yerevan, Armenia. Read more about our plans to establish an AI Institute. YerevaNN is a non-profit funded by research grants, academic partnerships and generous donations. Here are the easiest ways to support us.

Research

Machine learning for biomedical data

[2022] BARTSmiles: large-scale generative masked language models for molecular representations . [preprint] [code] (with IMB NAS RA, Toxometris.ai and Meta AI)
[2022] Improved molecular representations for property prediction based on VAEs. Published in Journal of Cheminformatics. (with Institute of Molecular Biology NAS RA).
[2020] Parasite: a method to align sentences in parallel corpora [code]. Helped to win WMT'20 Biomedical Translation Shared Task for en-ru pairs [paper].
[2019] BioRelEx: biological relation extraction dataset [data] [baselines] Published in BioNLP'19 workshop (with USC ISI and Bioinformatics group at IMB NAS RA, as part of the Big Mechanism program)
[2018] Clinical time series prediction benchmarks [code] [pretrained models]. Published in Nature Scientific Data (with USC ISI)

Machine learning under domain shift

[2023] Identifying and disentangling spurious features in pretrained image representations. [preprint] (with Meta AI and USC ISI). Accepted at ICML 2023 Workshop on Spurious Correlations, Invariance and Stability.
[2021] Characterization of the failure modes of domain generalization algorithms. Published in CVPR'22 (with USC ISI).
[2020] Robust classification under class-dependent domain shift [preprint] (with USC ISI). Presented at ICML 2020 UDL Workshop

Machine learning algorithms

[2023] Scaling laws for mixed-modal language models. [preprint] (with Meta AI)
[2023] Wireless non-line-of-sight localization of a device using a single antenna in an urban environment Accepted at IEEE Big Data Service 2023. [preprint] (with Yerevan State University and CNR Institute of Informatics and Telematics)
[2022] Matching map recovery with an unknown number of outliers. Published in AISTATS'23 [preprint] (with ENSAE-CREST)
[2022] GradSkip: an extension of a local gradient method for distributed optimization that supports variable number of local gradient steps in each communication round. [preprint] (with KAUST)
[2021] WARP: a parameter-efficient method for transfer learning in NLP. Published in ACL'21 (with USC ISI)
[2021] Theoretical analysis of the detection of the feature matching map in presence of outliers. Published in Electronic Journal of Statistics (with ENSAE-CREST)
[2021] A survey of deep neural networks for semi-supervised image classification. Published in JUCS.
[2020] A novel robust estimator of the mean of a multivariate Gaussian distribution. Published in Annals of Statistics (with ENSAE-CREST)
[2019] T-Corex: a novel method for temporal covariance estimation using information theoretic apparatus [preprint] [code] (with USC ISI)

Development of Armenian treebanks

Eastern Armenian treebank as part of the Universal Dependencies project [data (52K tokens)] [website] [code]
An end-to-end syntax parser for Armenian [code] [demo]
Collaboration with Marat Yavrumyan and Anna Danielyan from Yerevan State University. Partially funded by ANSEF, ISTC and Gulbenkian Foundation.

Team

Hrant Khachatrian
Tigran Galstyan
Ani Tevosyan
Khazhak Galstyan
Knarik Mheryan
Philipp Guevorguian
Rafayel Mkrtchyan
Menua Bedrosian

Karen Hambardzumyan
Ani Vanyan
Gayane Chilingaryan
Rafayel Darbinyan
Alla Barseghyan
Vahan Huroyan
Hakob Tamazyan
Tatev Vardanyan

Alumni: Hrayr Harutyunyan, Gor Arakelyan, Martin Mirakyan, Ashot Matevosyan, Arshak Minasyan, Hovhannes Tamoyan, Davit Papikyan, Arto Maranjyan Please fill in this form if you are interested in joining YerevaNN.

The handwritten digits in the background are generated by deep convolutional generative adversarial networks [paper] [code]