The year of 2023 marks the 14th year of Wang Xiaojing’s opening speech for the “Computational and Cognitive Neuroscience Summer School”. Standing at the podium of the academic center of the Cold Spring Harbor Asia, he briefly reviewed the developmental history of the emerging field of computational neuroscience. When he mentioned the classic paper by Hopfield, which used the spin glass model from physics to explain the brain’s associative memory system, he paused and asked the students if they had read it.

 

About half of the people in the classroom raised their hands. Wang Xiaojing nodded with satisfaction. Apparently, this exceeded his expectations.

 

More than 30 years ago, it was precisely Hopfield’s scientifically elegant research that captivated him. This prompted the theoretical physicist (a Ph.D. student then), to shift research focus to neuroscience. He borrowed the mindset and mathematical modeling used to study complex physical systems to help unravel the mysteries of the brain.

 

It was an arduous and off-beatean path, and Wang Xiaojing felt that he might not have persevered at that time was it not for an opportunity to attend a summer school in the United States. It was precisely for this reason that in 2010, he founded the first domestic Computational Neuroscience Summer School, aiming to pass on his unwavering commitment and passion for exploring the frontiers of science to the next generation of young scholars.

Wang Xiaojign is delivering a lecture at the 2023 Computational and Cognitive Neuroscience Summer School. Image Source: Cold Spring Harbor Asia

 

Outside the brightly-lit classroom, the glistening jade-blue ripples of Dushu Lake dissolved the summer heat of Suzhou. However, students’ enthusiasm in raising questions to Wang Xiaojing didn’t fade away with the scorching heat.

 

In order to uncover the story behind this summer school, the media team of NextQuestion visited the Cold Spring Harbor Asia in Suzhou twice, delving into the classrooms and living quarters. We have also conducted a video interview with this distinguished scholar and educator as the summer school drew to a close.

 

In his view, the summer school stood as a testimony to the rapid development of neuroscience in China over the past decade. Wang Xiaojing is deeply touched by the significant changes he has witnessed in Chinese students. Sometimes, he is amazed by the wealth of information they possess and the passion they have for exploration.

 

Computational neuroscience is a highly interdisciplinary field. For these young students who, like him years ago, find themselves at the crossroad of research but in an intensified age of involution, Wang Xiaojing shared his thoughts and experiences.

 

For the sake of readability, we have condensed the text for our readers’ enjoyment.

Question: Could you please begin by introducing your research field?

 

Professor Wang: I specialize in computational (theoretical) neuroscience, a relatively new field with a history of just over thirty years. As we all know, theory plays an unquestionably significant role in exploring the mysteries of the universe in physics. A similar situation is unfolding in the field of neuroscience: we increasingly need theoretical and mathematical modeling to understand how the brain works. The combination of experimental research and theoretical models can help us better unravel the mysteries of complex brain functions such as cognition, learning, and memory.

 

Question: What do you think are the most intriguing aspects of computational neuroscience?

 

Professor Wang: What makes computational neuroscience fascinating is its interdisciplinary nature. This means that people involved in this field come from various academic backgrounds, including mathematics, physics, computer science, and information science and engineering.

 

Those with a mathematical background can contribute to modeling and analysis, with dynamical systems theory playing a crucial role in understanding neural networks. Those specialized in information science may address problems from an information processing perspective. For example, brain-machine interfaces (BMI) need to make use of information processing methods to decode EEG signals and convert them into intructions that can be understood by machines. Those with a physics background may be inclined to view neuroscience as an open, complex dynamical system. This perspective means treating the brain as a self-organizing system, and studying the dynamics and functionality of neural network through the lens of the nonlinear behavior and complexity of the brain.

 

Question: What was the turning point for you to switch from pursuing a doctoral degree in physics to researching neuroscience back then?

 

Professor Wang: Sometimes, the factor of chance can alter the course of one’s life. I geared my research toward neuroscience in 1987. I believe there were two reasons for this shift. One was my doctoral studies in Brussels, Belgium, where Professor Prigogine, the director of the institute, had a remarkably broad range of interests. He used statistical physics and dynamical systems to study various systems, including biological ones. He played a role in kindling my interest in the brain.

 

Editor’s Note: Ilya Romanovich Prigogine (Russian: Илья́ Рома́нович Приго́жин) was a Belgian-Russian Jewish chemist and physicist. He was awarded the Nobel Prize in Chemistry in 1977 and is considered the founder of non-equilibrium statistical physics and the theory of dissipative structures. He expanded the application of the second law of thermodynamics, established by Clausius (R.J.E) almost a century earlier, to study thermodynamic phenomena in non-equilibrium states, opening up a new field that had previously received little attention. His work is regarded as one of the most significant advances in theoretical physics, theoretical chemistry, and theoretical biology in the past few decades.

 

The second factor was a model called “spin glass” which was used in the active physics study of complex systems. Glass appears to be quite simple, but it is not a simple crystal. Its structure is highly complex. In the 1980s, Hopfield used the physics theory of glass systems to study memory in the brain. His model at the time was very intriguing and drew the attention of many physicists. Since then, many researchers, including myself, began thinking about how to use the physics theory of complex systems to study the brain. Hopfield’s paper was also a major factor that motivated me to enter this field[1].

 

Editor’s Note: Spin Glass is a concept in physics commonly used to describe a particular phenomenon within complex systems. These systems typically consist of a large number of microscopic components (such as atoms or molecules) that possess spin properties, similar to the orientation of magnetic moments.

 

The formal beginning of computational neuroscience can be traced back to 1988. In that year, an article titled Computational Neuroscience was published in the journal Science, which can be considered as the manifesto of the field[2]. Also, in the same year, the first Computational Neuroscience Summer School was held at the Marine Biological Laboratory in Massachusetts. I was admitted to the inaugural summer school, which provided me with the opportunity to enter the field of computational neuroscience.

 

That was a very rewarding experience. Firstly, it helped me gain insights into the field of biology. While studying computational neuroscience, I gradually learnt about the mysteries of the brain. It became apparent that merely constructing abstract mathematical models was insufficient. I needed a deep understanding of neurobiology, including how an individual neuron functions, the functions of different brain regions, and the behaviors of different species. Secondly, during the summer school, I had the opportunity to meet some pioneers in the field of neuroscience. Thirdly, I engaged in discussions with both students and teachers, including debates, which unconsciously shaped the culture of my future research team.

 

Question: When did you first conceive the idea of creating your own Computational and Cognitive Neuroscience Summer School, and why did you choose China as the initial location?

 

Professor Wang: In 2010, I published an article on Science Times[3] introducing computational neuroscience. At that time, this field was still relatively new in the Asia-Pacific region. To provide an opportunity for more young people from mathematics, physics, and other disciplines to change the course of their research, it was necessary to offer a chance to learn about computational neuroscience. Reflecting on how I had benefited from the summer school at the Marine Biological Laboratory, I discussed with Professor Wu Si, who had recently returned to China, and others about establishing a similar international summer school in China. We decided to name it Computational and Cognitive Neuroscience (CCN).

 

At that time, computational neuroscience primarily focused on the brain systems related to primary sensations (such as vision) and motor behavior. Research on the neural mechanisms behind higher-level cognition began to develop only in this century. I happened to be working in this new field. So, I decided that CCN should emphasize cognition. This is the most significant difference between CCN and other summer schools in the field.

 

The other two original initiators were Zach Mainen and Upinder Bhalla. At that time, Cold Spring Harbor Laboratory had just established its Asian center in Suzhou, known as CSH-Asia, which provided an international platform. CCN became its first summer school. It attracted students from all over the world and formed a team of top-notch faculties. Such an international team allowed Chinese and overseas students to learn from and interact with each other. Over the past decade, many students who attended CCN have become leaders in the field, including Yang Guangyu, a tenure track assistant professor at MIT, Alan Antecovic, a tenured associate professor at Yale University, Li Songting, a full professor at Shanghai Jiao Tong University, Jean-Rémi King, a researcher at Meta, Francis Song, a researcher at OpenAI, and many others. The success of CCN is the result of collective effort, and I am very grateful for the contributions of Wu Si, Zach Mainen, Upi Bhalla, Li Songting, Huang Chengcheng, Yang Guangyu, Dora Angelaki, Christopher Honey, and the teaching assistants. My gratitude also goes to the support of the organizing institutions.

 

Question: What were your initial thoughts and ideas when founding the summer school? What are the distinctive features of the summer school?

 

Professor Wang: The primary goal was to help young researchers enter the field of neuroscience and learn how to raise interesting questions of scientific value.

 

Firstly, in the field of neuroscience, choosing a research direction is crucial. For example, if I want to understand the neural mechanisms of consciousness but didn’t have specific hypotheses and methods for validating those hypotheses, it would be empty talk.

 

Historically, neuroscience mainly focused on perception and motor function, which could be studied using relatively simple animal models like frogs or fruit flies. However, for more complex functions like cognition, it was necessary to focus on challenging yet specific questions that could be addressed with experimental and theoretical approaches.

 

Secondly, students will receive comprehensive education from various disciplines. The mentors of the summer school come from fields such as biology, physics, mathematics, and more, including experts in molecular and cellular biology, network systems, and behavioral neuroscience. Students have the rare opportunity to learn about different research paradigms while expanding their research horizons, which has a significant impact on their future growth.

 

Thirdly, we emphasize the cultivation of an academic culture. During the summer school, we invite scientists from both China and abroad not only to give lectures but also to engage in in-depth discussions and interactions with students. They spend a considerable amount of time on the campus, dining with students, and participating in extracurricular activities. Students have the opportunity to engage in discussions with their professors on an equal footing, gradually becoming comfortable expressing their own views and daring to challenge academic authorities, even when faced with academicians and renowned scientists. Additionally, getting to know scientists from around the world provides students with more choices and opportunities in their future academic paths.

 

Question: What changes have you observed in China’s training of interdisciplinary talent over the past decade or so?

 

Professor Wang: Compared to over a decade ago, students today may already have more knowledge and maturer thoughts before entering the field of neuroscience. Furthermore, they have a strong curiosity and a thirst for knowledge. They are enthusiastic about understanding new developments and cutting-edge advances, and I find these two traits very valuable.

 

 

Additionally, one hot topic of concern for young people today is AI, but the road ahead is still long before AI system can catch up with the human brain. Our understanding of the intelligent biological mechanisms of humans remains insufficient and fundamental research requires long-term efforts. Some people are passionate about pioneering technological breakthroughs while others focus more on basic research; and some are enthusiastic about using neuroscience to advance psychiatry. Young people are pursuing what they truly love for a lifetimat, and each one of them have a different answer. If you’re researching a topic that’s genuinely interesting for you, you can take your time instead of following suit.

 

Question: In summary, what are the characteristics and challenges of training talent in computational neuroscience?

 

Professor Wang: Firstly, it’s interdisciplinary. Students with backgrounds in mathematics or physics can learn about brain biology and understand experimental neuroscience research during the summer school. Conversely, individuals conducting biological experiments often lack a strong mathematical background, and the summer school helps familiarize them with mathematical models. For example, in this summer school, there was a psychology student who remarked that he learned many methods for mathematical modeling.

 

Secondly, it’s about institutions. In China, there are relatively few institutions that support interdisciplinary research and education. The summer school has been quite unique and successful for over a decade, and there is a consensus on the necessity of developing computational neuroscience and talent. However, long-term funding support is not always guaranteed.

 

I would like to express special gratitude to Mr. Chen Tianqiao and his wife Chrissy Luo for founding the Tianqiao and Chrissy Chen Institute (TCCI), as well as the support from the James Simons Foundation and DeepMind. Their support has been crucial.

 

Question: How can the summer school address some shortcomings in regular university education?

 

Professor Wang: In general, universities often have very specialized disciplines, which may not be conducive to interdisciplinary research and the cultivation of interdisciplinary young talent. When we started New York University Shanghai from scratch back then, as Provost, I took the opportunity to plan and implement a new model: research would not be conducted based on departments but revolve around several interdisciplinary centers. Similarly, summer schools like CCN are entirely interdisciplinary and serve a different purpose than regular university education. We aim at attracting talent from various fields to work together on advancing neuroscience, which deserves attention and support.

 

Interview by Yu Hanqi
Edited by Xu Yunke & Wang Ruyue