Follow our latest advances in machine learning.

Distributional shift

Generative models

Graph machine learning

Nearest neighbor search

Uncertainty estimation 

Computer vision

Machine learning theory

Natural language processing 

Tabular data

Bayesian methods

Gradient boosting

Large-scale machine learning

Blog

Yandex Research Blog

Nearest neighbor search is a long-standing problem arising in a large number of machine learning applications, such as recommender services, information retrieval, and others.

This competition focuses on developing indexing data structures and search algorithms for practical settings of the Approximate Nearest Neighbor (ANN) or Vector search problem. More details are on the competition’s website. We encourage everyone to join by August 30th, 2023!


NeurIPS 2023 Competition Track: Big-ANN

A paper has been accepted for publication at the International Conference on Computer Vision (ICCV 2023). DeDrift: Robust Similarity Search under Content Drift by Dmitry Baranchuk, Matthijs Douse, Yash Upadhyay, Zeki Yalniz.

Paper accepted to ICCV 2023

At the International Conference on Machine Learning (ICML 2023), Yandex Research will present papers on modelling tabular data with diffusion models, a model-parallel training algorithm designed for poorly connected devices and other topics. We also co-author a tutorial ​​on reinforcement learning from human feedback.

Yandex at ICML 2023

Today, training most powerful models often takes significant resources. Our research aims to make large-scale training more efficient and accessible to the entire machine learning community.

Language is one of the key forms of communication. We study methods of language representation and understanding to simplify human-computer interactions.

Large language models are among the most significant recent advances in machine learning. Still, leveraging these models can be difficult: offloading and quantization have limitations, and third-party APIs are less flexible. We propose Petals, an open-source decentralized system (showcased this week at the ACL 2023 Demonstrations track) allowing anybody to run large models or even adapt them using the idle resources of volunteers. In this post, you will learn the motivation behind the system, its underlying ideas, and its advantages compared to other ways of using large models.

Petals: decentralized inference and finetuning of large language models

Graphs are a natural way to represent data from various domains such as social networks, molecules, text, code, etc. We develop and analyze algorithms for graph-structured data.

Distributional shift is the mismatch between training and deployment data that is ubiquitous in the real-world. Studying this phenomenon can enable safer and more reliable ML systems.

Uncertainty estimation enables detecting when ML models make mistakes. This is of critical importance in high risk machine learning applications, such as autonomous vehicle and medical ML.

We propose a general approach for inducing diverse distributional shifts based on graph structure. In particular, we compute an arbitrary node-level characteristic that reflects some structural property of nodes. Then, we sort all nodes in ascending order of this characteristic and split them into different subsets — those with the smallest values are considered to be in-distribution (ID), while the remaining ones are out-of-distribution (OOD). As a result, we obtain a graph-based distributional shift where ID and OOD nodes have different structural properties.

Blog

Posts

NeurIPS 2023 Competition Track: Big-ANN

Paper accepted to ICCV 2023

Yandex at ICML 2023

Petals: decentralized inference and finetuning of large language models

Structural distributional shifts for graph datasets

Filter by: