🎉 Time to celebrate!
Milena Gazdieva, a research engineer at the Skoltech Artificial Intelligence Center, has won the Yandex ML Prize award in the category for early-career faculty members!
At Skoltech, Milena teaches a course on generative AI, and her research interests lie in generative models based on optimal transport theory. A project based on one of these models was developed by participants of the 2024 Skoltech SMILES Summer School. The team won the award for best project among online participants.
The course developed with her participation covers both classical topics (Gaussian mixture models, variational autoencoders, etc.) and modern flow-based models — from fundamental principles to current research directions.
The Yandex ML Prize has been awarded annually since 2019 and aims to support teachers and researchers in the field of AI and machine learning. The winners receive monetary awards, grants to use the Yandex Cloud platform, and access to Yandex 360 services.
👏 Congratulations to Milena, and we wish her inspiration for new scientific and educational achievements!
Milena Gazdieva, a research engineer at the Skoltech Artificial Intelligence Center, has won the Yandex ML Prize award in the category for early-career faculty members!
At Skoltech, Milena teaches a course on generative AI, and her research interests lie in generative models based on optimal transport theory. A project based on one of these models was developed by participants of the 2024 Skoltech SMILES Summer School. The team won the award for best project among online participants.
"I am very happy that the award committee valued my teaching activities and my contribution to the development of courses on generative AI so highly. It all started with a mini-course that I helped develop and deliver at the Yandex School of Data Analysis in 2023. Then, in 2024, with my participation, this course was gradually refined and improved: in different versions, it was taught at several institutions and summer schools — for example, at the Skoltech SMILES Machine Learning Summer School. Finally, in the spring of 2025, we developed the course into Skoltech's first large-scale course on generative AI," said Milena.
The course developed with her participation covers both classical topics (Gaussian mixture models, variational autoencoders, etc.) and modern flow-based models — from fundamental principles to current research directions.
The Yandex ML Prize has been awarded annually since 2019 and aims to support teachers and researchers in the field of AI and machine learning. The winners receive monetary awards, grants to use the Yandex Cloud platform, and access to Yandex 360 services.
👏 Congratulations to Milena, and we wish her inspiration for new scientific and educational achievements!
🔥8👍3❤1❤🔥1👏1
👏 Distinguished Professor Artem R. Oganov named among the world’s top researchers for the third time
Clarivate reports that only one in a thousand researchers is included in this ranking. To qualify, researchers must have several publications ranked in the top 1% in their field according to citations in the Web of Science over the past 11 years. They must also comply with several other strict criteria based on a thorough analysis and expert assessments by the U.S. Institute for Scientific Information.
On November 14, Artem and his colleagues presented USPEX 25, a new version of the algorithm combining machine learning and fundamental physics. USPEX 25 can be used for cutting-edge research on a standard computer, even without access to a supercomputer.
Clarivate reports that only one in a thousand researchers is included in this ranking. To qualify, researchers must have several publications ranked in the top 1% in their field according to citations in the Web of Science over the past 11 years. They must also comply with several other strict criteria based on a thorough analysis and expert assessments by the U.S. Institute for Scientific Information.
Artem R. Oganov: “2025 has been a very successful year for me and my lab. This ranking recognizes all my students and colleagues as well. I am very happy doing the work I love in my own country and collaborating with such a brilliant team.”
On November 14, Artem and his colleagues presented USPEX 25, a new version of the algorithm combining machine learning and fundamental physics. USPEX 25 can be used for cutting-edge research on a standard computer, even without access to a supercomputer.
🔥10❤4👍3
⭐️ The Industrial Immersion Excellence Award was presented to the MSc students for their summer work on real-world tasks at one of our partner companies — we will cover each of these stories separately.
We are opening the series with the internship recognized as the best from the "Advanced Manufacturing Technologies" Master's program. Gleb Zhdanko completed his internship at B&B Industries, where he worked on calculating the power output of a wind turbine with custom blades.
The student noted that the simulation results showed an error of less than 5% within the turbine's operational range, confirming the high reliability of the calculations.
#IndustrialImmersion
We are opening the series with the internship recognized as the best from the "Advanced Manufacturing Technologies" Master's program. Gleb Zhdanko completed his internship at B&B Industries, where he worked on calculating the power output of a wind turbine with custom blades.
"For the calculations, I used OpenFAST—an open-source software package for the numerical simulation of turbines," Gleb explained. "This approach significantly accelerates the development of new wind turbines by allowing for rapid assessment of how changes in blade profiles affect turbine efficiency. As a result, the number of necessary, costly full-scale tests is substantially reduced."
The student noted that the simulation results showed an error of less than 5% within the turbine's operational range, confirming the high reliability of the calculations.
#IndustrialImmersion
❤7👍2
🎉 Time to celebrate!
Junior Research Scientist Dmitrii Kryukov from the Bio Center of Skoltech received the Best Pitch Award at the 3rd Healthy Longevity Symposium, which was dedicated to the challenges of aging.
The symposium, held in Abu Dhabi, was hosted by Khalifa University and organized in partnership with the Nature Conferences’ "Measuring Healthy Aging" event.
At the symposium, Dmitrii presented a poster on developing a methodology for biological age prediction using Large Health Models (LHM) — artificial intelligence trained on large volumes of medical data.
We wish Dmitrii continued success in his future research and hope his work earns well-deserved recognition from colleagues worldwide! 👏
Junior Research Scientist Dmitrii Kryukov from the Bio Center of Skoltech received the Best Pitch Award at the 3rd Healthy Longevity Symposium, which was dedicated to the challenges of aging.
The symposium, held in Abu Dhabi, was hosted by Khalifa University and organized in partnership with the Nature Conferences’ "Measuring Healthy Aging" event.
At the symposium, Dmitrii presented a poster on developing a methodology for biological age prediction using Large Health Models (LHM) — artificial intelligence trained on large volumes of medical data.
We wish Dmitrii continued success in his future research and hope his work earns well-deserved recognition from colleagues worldwide! 👏
🔥11🎉3👍2❤1
⭐️ A new week brings a new story about one of the best industrial immersions of Skoltech students.
Our series of publications continues with a feature on the top internship project in the Next Generation Communication Technologies master’s program. The author, Alexander Shmatok, completed his internship at Heliosoft, where he worked on predicting user error vector magnitude in signals transmitted from 5G base stations operating in multi-user mode.
He noted that by applying mathematical theorems, he managed to establish a relationship between user data and the error rate, successfully predicting it with an accuracy of over 80% (a margin of error no greater than 20%).
#IndustrialImmersion
Our series of publications continues with a feature on the top internship project in the Next Generation Communication Technologies master’s program. The author, Alexander Shmatok, completed his internship at Heliosoft, where he worked on predicting user error vector magnitude in signals transmitted from 5G base stations operating in multi-user mode.
"When there are many users, one user's signal can interfere the signal for another," Alexander explained. "For a base station to operate efficiently, it's necessary to predict the error vector magnitude for each user based on their location and signal power. This is crucial for optimizing the selection of error-correcting codes."
He noted that by applying mathematical theorems, he managed to establish a relationship between user data and the error rate, successfully predicting it with an accuracy of over 80% (a margin of error no greater than 20%).
#IndustrialImmersion
👍5❤1
📸 The Skoltech Student Council organized an excursion to the Joint Institute for Nuclear Research (JINR) in Dubna.
During the visit, master's and phd students immersed themselves in the Institute's scientific life: they learned about educational programs and cutting-edge research, and visited an interactive exhibition of JINR projects.
The students were particularly impressed by the tour of the NICA collider, where they examined in detail the Multi-Purpose Detector (MPD), designed for experiments with heavy-ion collisions.
We thank the engineers and staff of JINR for a fascinating and insightful day!
During the visit, master's and phd students immersed themselves in the Institute's scientific life: they learned about educational programs and cutting-edge research, and visited an interactive exhibition of JINR projects.
The students were particularly impressed by the tour of the NICA collider, where they examined in detail the Multi-Purpose Detector (MPD), designed for experiments with heavy-ion collisions.
We thank the engineers and staff of JINR for a fascinating and insightful day!
❤11
🔋 Researchers from Skoltech, MIPT, and the RAS Institute of Nanotechnology of Microelectronics have achieved a fivefold increase in the capacitance of carbon nanowalls, a material used in the electrodes of supercapacitors. These are auxiliary energy storage devices used in tandem with conventional batteries in electric vehicles, trains, port cranes, and other systems. The properties of carbon nanowalls were enhanced by irradiating the material with an optimal dose of argon ions. The study was published in Scientific Reports and supported by the Russian Science Foundation.
Unlike conventional energy storage, supercapacitors can accumulate or release energy almost instantaneously. This makes them an excellent complement to electrochemical storage during peak energy demand — such as during power grid surges, cargo lifting, or the initial acceleration of electric transport. Furthermore, an electric train equipped with supercapacitors can recuperate energy during braking. They also operate over a wider temperature range, are less prone to wearing out or catching fire.
The material itself can be imagined as vertically oriented graphene layers. This branched structure results in large specific surface area, which ultimately translates into higher specific capacitance for the current source. To further improve the characteristics, an argon ion implantation procedure was performed, creating beneficial defects in the structure.
This advancement paves the way for efficient energy storage devices that can adapt to peak loads and extend the lifespan of primary batteries, with potential future applications ranging from transportation to power grids.
Unlike conventional energy storage, supercapacitors can accumulate or release energy almost instantaneously. This makes them an excellent complement to electrochemical storage during peak energy demand — such as during power grid surges, cargo lifting, or the initial acceleration of electric transport. Furthermore, an electric train equipped with supercapacitors can recuperate energy during braking. They also operate over a wider temperature range, are less prone to wearing out or catching fire.
“The more energy supercapacitors can store, the more applications they will find. We are investigating ways to improve their characteristics by various treatments of the carbon material used in their electrodes,” says the study’s principal investigator, Assistant Professor Stanislav Evlashin of Skoltech Materials. “Earlier this year we showed that capacitance can be enhanced by incorporating atoms of other elements into carbon nanowalls. This time we have achieved a more pronounced increase in capacitance by treating that carbon-based material with argon ions at an ion accelerator. We determined which dose of ions is optimal for maximizing useful defects without damaging the material too much.”
The material itself can be imagined as vertically oriented graphene layers. This branched structure results in large specific surface area, which ultimately translates into higher specific capacitance for the current source. To further improve the characteristics, an argon ion implantation procedure was performed, creating beneficial defects in the structure.
This advancement paves the way for efficient energy storage devices that can adapt to peak loads and extend the lifespan of primary batteries, with potential future applications ranging from transportation to power grids.
❤2👍1
🌍 From Suzhou to Geneva: Skoltech scientists present their research on the international stage.
📍 China
— In Suzhou, the team from the AI Center’s Natural Language Processing Laboratory gave presentations at the 30th Conference on Empirical Methods in Natural Language Processing (EMNLP, A-star). Associate Professor Alexander Panchenko, who heads the laboratory, Senior Research Scientist Vasily Konovalov, Research Engineers Sergey Pletenev and Mikhail Salnikov, as well as PhD student Maria Marina ("Computational and Data Science and Engineering" program) and PhD alumnus Artem Vazhentsev presented their research on improving language models, detecting AI hallucinations, and automating clinical coding.
— In Nanjing, Faridun Jalolov, a Research Intern at the Energy Transition Center and a PhD student in the Materials Science program, along with Michel Lukanov, a PhD student in the Materials Science program, visited the Nanjing University’s Research Group in Physics and Materials at Extremes, led by Professor Jian Sun. The visit was part of a joint project on developing active learning methods for simulating complex materials, led by Aleksandr Kvashnin, a Professor at the Energy Transition Center.
📍 United Arab Emirates
— In Abu Dhabi, Dmitry Kriukov, a Junior Research Scientist at the Center for Molecular and Cellular Biology, received the award for best pitch at the Nature Forum: Measuring Healthy Aging, held as part of a symposium at Khalifa University. Dmitry presented a new method for calculating biological age using large health models — AI systems trained on extensive medical data.
📍 Austria
— In Vienna, Associate Professor Dzmitry Tsetserukou, Head of the Engineering Center’s Intelligent Space Robotics Laboratory, delivered a talk and moderated a session at the 3rd International Conference on Foundation and Large Language Models (IEEE FLLM 2025). The conference featured a joint research project presenting GestOS — a gesture-based operating system for controlling heterogeneous robot teams.
📍 Switzerland
— In Geneva, Associate Professor Ivan Sergeichev, who heads the Materials Center, participated in the 67th session of the UN Subcommittee on the Transport of Dangerous Goods as chairman of a working group. Alexey Fedorenko, a Senior Research Scientist at the Center, also joined the group. The collaborative work is aimed at developing international standards for the safe transportation of liquefied gases.
📍 China
— In Suzhou, the team from the AI Center’s Natural Language Processing Laboratory gave presentations at the 30th Conference on Empirical Methods in Natural Language Processing (EMNLP, A-star). Associate Professor Alexander Panchenko, who heads the laboratory, Senior Research Scientist Vasily Konovalov, Research Engineers Sergey Pletenev and Mikhail Salnikov, as well as PhD student Maria Marina ("Computational and Data Science and Engineering" program) and PhD alumnus Artem Vazhentsev presented their research on improving language models, detecting AI hallucinations, and automating clinical coding.
— In Nanjing, Faridun Jalolov, a Research Intern at the Energy Transition Center and a PhD student in the Materials Science program, along with Michel Lukanov, a PhD student in the Materials Science program, visited the Nanjing University’s Research Group in Physics and Materials at Extremes, led by Professor Jian Sun. The visit was part of a joint project on developing active learning methods for simulating complex materials, led by Aleksandr Kvashnin, a Professor at the Energy Transition Center.
📍 United Arab Emirates
— In Abu Dhabi, Dmitry Kriukov, a Junior Research Scientist at the Center for Molecular and Cellular Biology, received the award for best pitch at the Nature Forum: Measuring Healthy Aging, held as part of a symposium at Khalifa University. Dmitry presented a new method for calculating biological age using large health models — AI systems trained on extensive medical data.
📍 Austria
— In Vienna, Associate Professor Dzmitry Tsetserukou, Head of the Engineering Center’s Intelligent Space Robotics Laboratory, delivered a talk and moderated a session at the 3rd International Conference on Foundation and Large Language Models (IEEE FLLM 2025). The conference featured a joint research project presenting GestOS — a gesture-based operating system for controlling heterogeneous robot teams.
📍 Switzerland
— In Geneva, Associate Professor Ivan Sergeichev, who heads the Materials Center, participated in the 67th session of the UN Subcommittee on the Transport of Dangerous Goods as chairman of a working group. Alexey Fedorenko, a Senior Research Scientist at the Center, also joined the group. The collaborative work is aimed at developing international standards for the safe transportation of liquefied gases.
🔥6👍2
⚡️ Scientists from Skoltech, Sirius University, and Zhejiang University have proposed a new approach for analyzing and optimizing the long-term operation of vanadium flow batteries across a wide range of ambient temperatures. The developed model, validated by experiments, allows for accurate prediction of system efficiency and energy capacity, which is critically important for their integration into power grids and for use in industrial energy storage systems within Russia’s climatic conditions. The research has been published in the Applied Energy journal.
Vanadium flow batteries are among the top contenders for the role of energy storage systems. Their key advantages are long service life, the ability to fully discharge without degradation, and flexible scalability to meet the needs of power systems. One of the main performance indicators of such batteries is their energy efficiency — the ratio of energy delivered during discharge to the energy spent on charging.
Thermal processes inside the battery pose a particular challenge. Large-scale energy storage systems are often installed outdoors, making them vulnerable to seasonal temperature fluctuations. This directly affects the viscosity of the electrolyte, the power of the pumps, and the overall stability of the system.
The development paves the way for creating “climate-adaptive” energy storage systems that can operate stably and cost-effectively across a wide temperature range, enhancing the reliability of green energy in Russia.
Vanadium flow batteries are among the top contenders for the role of energy storage systems. Their key advantages are long service life, the ability to fully discharge without degradation, and flexible scalability to meet the needs of power systems. One of the main performance indicators of such batteries is their energy efficiency — the ratio of energy delivered during discharge to the energy spent on charging.
Thermal processes inside the battery pose a particular challenge. Large-scale energy storage systems are often installed outdoors, making them vulnerable to seasonal temperature fluctuations. This directly affects the viscosity of the electrolyte, the power of the pumps, and the overall stability of the system.
“In the course of the research, an analytical formula for determining the equilibrium temperature of the battery, which establishes itself during long-term operation, was derived for the first time. This tool allows for a quick assessment of the system’s thermal state and prevents overheating without the use of complex computations. The model was validated on a 5 kW experimental setup and showed high accuracy: The calculation error for voltage was less than 2%, and for temperature — less than 3%,” commented the study’s lead author, Stanislav Bogdanov, a junior research scientist at the Skoltech Energy Center.
“The presented method serves as a tool for informed decision-making. For example, maximum power is the priority for emergency grid startup, while maximum efficiency is key for daily peak shaving. Our research provides practical recommendations for selecting optimal operating modes for industrial energy storage systems based on vanadium flow batteries,” noted Mikhail Pugach, a senior research scientist at the Skoltech Energy Center, co-author, and leader of the project and the study.
The development paves the way for creating “climate-adaptive” energy storage systems that can operate stably and cost-effectively across a wide temperature range, enhancing the reliability of green energy in Russia.
❤5
We look forward to seeing you until December 7 at Gostiny Dvor for the 27th International Fair of Intellectual Literature, non/fiction.
At the Paulsen Publishers booth (B-17), we are presenting the updated edition of the book Life and Other Stories: How We and the Living World Around Us Work, published by Skoltech and the Russian Science Foundation. This collection features interviews with leading biologists.
This year, the project team has also released an English-language edition and completed the project with conversations with the book’s protagonists and reports from remote corners of Russia on the updated landing page.
Please open Telegram to view this post
VIEW IN TELEGRAM
👍5❤2