The Up Great technology contest READ//ABLE is designed to stimulate the development of new approaches to machine learning to let the artificial intelligence to understand deeply the meaning of the text and to analyze the causal relationships in a wide range of topics.

In 2020 our team 'NeuroReading' won the 1st cycle Up Great technology contest READ//ABLE of the nomination "Grammar.Eng" for texts in English. Team demonstrated the highest quality solution (accuracy level – 84.5%).

In 2021 our team 'DeepPavlov' won the 2nd cycle Up Great technology contest READ//ABLE for texts in English. Team demonstrated the highest quality solution (accuracy level – 107%)

The DREAM Team was selected to compete for the Alexa Prize Socialbot Grand Challenge 4. This selection comes with $250K research grant, Alexa-ended devices, and AWS support to advance open-domain Conversational AI.

In March 2021, our researcher Leonid Pugachev and our interns Alim Adelshin took the first place in the competition Dialogue: Russian News Clustering and Headline Selection — Clustering. They developed a model for clustering news for Russian language, which showed a 96.85% accuracy as a result.

In December 2020, our researcher and the captain of Dream Alexa Prize team Dilyara Baimurzina took third place in the AI4 Humanities: ruGPT-3 competition from AIJourney. She developed an AI CopyWriter model for paraphrasing and / or distributing text based on ruGPT3Large.

In March 2020, our researcher Alexey Sorokin took second place in the GramEval-2020 competition. He developed a model for the automatic determination of the morphological and syntactic characteristics of words in sentences in Russian language.

The DREAM Team was selected by Amazon for Alexa Prize Socialbot Grand Challenge 3. DREAM Team has reached the semifinals of the Alexa Prize 3 and DREAM Socialbot achieved final average rating of 3.39.

DeepPavlov Team participated in Schema-Guided State Tracking track of the 8th Dialogue System Technology Challenge 8 (DSTC).

Since 2015, we held Six International week-long Scientific Hackathon, which were attended by about 1000 people, and 40 leading researchers in the field of machine intelligence.

In 2018 and 2019 we held International Conversational Intelligence Summer School (CISS 1&2), which were attended by about 100 people.

The Second Edition of NIPS Conversational Intelligence Challenge was a variation of the famous Turing test. During the competition people were randomly connected with other people or bots and made a conversation about their hobbies, work, family, etc. The aim of the competition was to compare and develop state-of-the-art methods for non-goal-oriented dialogue systems (chit-chat models). The task was, conditioned on a persona, to chit-chat to get to know the other speaker.

The organizing team comes from multiple groups — Moscow Institute of Physics and Technology (as known as DeepPavlov.ai), Facebook AI Research, Université de Montréa, McGill, and Carnegie Mellon University.

In 2017, the Conference on Neural Information Processing Systems (NIPS) was sponsoring an open competition called The Conversational Intelligence Challenge aimed at creating a chatbot that can hold an intelligent conversation with a human partner. The event produced evaluations of state-of-the-art dialogue systems and an open-source data set for the future training of end-to-end systems.

The competition was organized by Moscow Institute of Physics and Technology (as known as DeepPavlov.ai), Université de Montréal, McGill University, and Carnegie Mellon University in partnership with Facebook, Flint Capital, IVADO, Maluuba, and Element AI.