For Students

Information for prospective students

This page summarizes the characteristics of our lab and answers questions about lab assignment for those who are considering joining us. For the topics we work on, see Research Themes; for past bachelor's and master's thesis topics, see Members; and for the resulting outcomes, see Papers and Conference Presentations. For the Technical Seminar, see AI Seminar.

What We Value in the Lab

Thinking Things Through

The most important thing in carrying out research is to think.

When setting a research theme, you need to think about what specific theme to pursue based on a great deal of information. And once the theme is set, you need to think about how to achieve your goal. Thinking everything through on your own can be hard. In that case, it is a good idea to ask others, such as faculty members and lab members, for their views. Integrating your own thinking with that of others gives rise to new ideas.

Taking On Challenges

It would be a waste not to take on a challenge when the opportunity is there.

Once your thinking has taken shape to some extent, go ahead and give it a try. Knowledge and skills gained through your own experience are easier to retain. Moreover, sometimes the opportunity to take on a challenge simply is not there in the first place. Having such an opportunity is itself a stroke of good fortune, so take on challenges actively.

Not Being Afraid of Failure

You gain far more from failure than you do from success.

When you fail, thinking about questions such as "Why did it fail?" and "What should I do to succeed?" can lead you to success. Also, actively take on challenges even when they look likely to fail. Sometimes what looked likely to fail does not fail at all and becomes the invention of the century. The blue LED now used in lighting and displays was invented thanks to malfunctioning experimental equipment. Had the experiment been called off because of that malfunction, I do not believe this invention of the century would ever have come about. For these reasons, our lab places importance on thinking, on actively taking on challenges without fear of failure, and on learning from failure when it happens. These habits are useful beyond research as well. Once you graduate from university, it becomes harder to fail. While you are in an environment where you can afford to fail, master how to learn from failure. That is why we emphasize taking on challenges actively, and thinking in order to earn opportunities to take them on.

Frequently Asked Questions

Lab assignment is decided through a comprehensive evaluation of your motivation for applying, what you have learned at university, and how well you match the lab. Regarding your university studies, grades do matter, but we place even greater weight on questions such as "Why did you take that course?" and "What did you learn from taking it?"

There are no required courses for assignment to the lab. This is because the techniques and skills you will need are not fixed until your specific research theme is decided. Of course, there are courses, such as systems engineering, programming, and AI, that can be put to use in many research projects, but they are not essential for every project. Research proceeds by making full use of the knowledge and skills you have already learned or will newly learn. Moreover, knowledge and skills that seem unrelated to research at first glance can sometimes be applied to it through a sudden flash of insight. Therefore, there are no specific courses you must take. It is good if you have consolidated your knowledge and skills so that you can apply what you have learned so far.

We do not require any particular skills before you join the lab. We hope that you will actively acquire the skills that become necessary as your research progresses. That said, it helps if you are not averse to programming, algorithms, or using AI. Programming and AI allow you to handle many routine tasks efficiently.

Not a problem at all. In fact, we believe that being able to use AI as a tool will be essential in the years ahead.

Our lab runs the AI Seminar so that assigned students can acquire the latest AI technologies. In this seminar, you can learn step by step, from the fundamentals of generative AI to building AI agents.

Even if your research theme itself is not AI, we have a support system in place so that you can use AI as a powerful partner to accelerate your research. As long as you are interested in AI and motivated to use it to try something new, no prior knowledge is required.

For the broad theme concerning society, we would like you to choose something you are interested in, for example, "I want to build an artificial society" or "I want to do research that saves lives." When narrowing a broad theme down to a specific one, we decide together with you through literature surveys and seminars. Of course, we can also propose themes from our side.

For students who intend to go on to the master's program, we discuss and set a three-year research plan immediately after lab assignment, taking into account their intended career path after completing the master's degree (such as the type of job or industry), so that they can present their work effectively in internships and job hunting.

During the teaching period, seminars are held in person. During vacation periods, they are held online as needed. No seminars are planned during the New Year holidays, the Obon holidays, and similar breaks.

Below is an example of the general annual schedule for students who are assigned to the lab from their fourth year. Note that starting in academic year 2026, lab assignment will take place in the fall semester of the third year, so the schedule will differ from the one below.

Fourth Year

Month	Event
Late February	Lab assignment decided
March	AI Seminar I¹
April–May	Literature survey; research theme decided
June	Writing the abstract; presentation practice
July	Comprehensive Research I presentation
August	AI Seminar II²; start of research work
September	Conference presentation³; implementation, etc.
October–November	Experiments, analysis, etc.
December	Thesis writing
January	Presentation practice
February	Comprehensive Research II presentation
March	Conference presentation⁴; graduation ceremony

A seminar for acquiring basic knowledge and skills in generative AI, which also serves as a first meeting for the students newly assigned to the lab. ↩
A seminar for acquiring practical knowledge and skills so that students can apply generative AI to advance their research. It is scheduled to be held after the Comprehensive Research I presentation, but the timing may shift from year to year. ↩
At the Comprehensive Research I presentation, students present to faculty members from different fields. At this conference presentation, by contrast, researchers and students in the same specialty gather in a retreat format to give and attend presentations, which provides an opportunity to obtain important hints and feedback for advancing research. It is usually held at Miura Kaigan. Travel and accommodation expenses are covered by the lab. ↩
Each year, students present the work they carried out in Comprehensive Research before researchers and students in the same field, as the culmination of their research. Those who wish to may present on site (it is often held on Ishigaki Island). Travel and accommodation expenses are covered by the lab. ↩

We hold seminars with all of the lab's students, but we do not have core hours in the sense of gathering at a fixed time every day.

We have no plans to standardize on a single programming language in the lab. We would like you to choose a language you are good at or one that makes your research easier to carry out, for example, Python for data analysis and machine learning, or CUDA C++ for developing GPU-based technologies. Harada himself uses a variety of languages depending on the purpose, including C++, C#, Rust, Python, and SQL. Please feel free to consult us about programming languages as you carry out your research. In recent years, students have typically carried out their research using languages such as C, C++, Java, and Python.

Not a problem at all. The lab maintains the following computing resources, which are configured for remote access as needed.

AMD EPYC 7743 × 2 (128 cores), 1024 GB ECC memory × 1 machine
AMD Ryzen Threadripper 3990x (64 cores), 128 GB ECC memory × 1 machine
AMD Ryzen Threadripper 1950x (16 cores), 128 GB ECC memory × 1 machine
AMD Ryzen 9 5950x (16 cores), 128 GB memory × 1 machine
AMD Ryzen 9 3950x (16 cores), 128 GB ECC memory × 6 machines
NVIDIA DGX Spark (128GB RAM) × 1 machine

In addition, in recent years GPUs have been used to accelerate general-purpose computing such as data analysis and machine learning. In particular, we have recently set up an environment for running and training large language models. The lab operates the following GPUs installed in the machines listed above.

NVIDIA RTX Pro 6000 Blackwell (96GB VRAM) × 4
NVIDIA GeForce RTX 4090 (24GB VRAM) × 1
NVIDIA RTX A6000 (48GB VRAM) × 2
NVIDIA GeForce RTX 3090 (24GB VRAM) × 1
NVIDIA GeForce TITAN RTX (24GB VRAM) × 2
NVIDIA GeForce RTX 2080 Ti (11GB VRAM) × 1

We cluster these resources to build a virtualization platform. This allows us to set up and provide not only Windows but also Linux environments such as Ubuntu and execution environments similar to Google Colab.