Getting Involved
- [Jan 2025] Actively hiring interns and MS/PhD students
- Sophomores and juniors are encouraged to apply
- Qualifications:
About RLLAB
The goal of our lab is to enable machines to learn complex real-world tasks, especially with robots. Toward this goal, we are interested in developing scalable reinforcement learning algorithms that can leverage diverse data and prior knowledge.
- Relevant topics: reinforcement learning, imitation learning, representation learning, generative modeling, and their applications to robotics
- Target conferences:
(Robot Learning)
RSS,
CoRL /
(Machine Learning)
ICLR,
ICML,
NeurIPS
We are looking for students who
- are motivated to make progress in AI, improve yourself, and build a stellar career
- get things done while having high standard on quality
- can work in-person and full-time for at least 6 months
- 40+ hours/week during summer and winter; 20+ hours/week otherwise
What to expect from our lab?
I aim to build a strong research lab, where everyone aims for a career at top schools and companies. Thus, my expectation for PhD is to match or even outperform students from other top labs in Korea (e.g. Prof. Joseph Lim’s lab, Prof. Jinwoo Shin’s lab).
By the time you graduate, I believe you will
- publish 1+ top-tier conference paper every year
- have good presentation skills
- mentor several junior students
- have 1-2 internship experience at top research labs
- be well-known in the community
To support your growth and success in PhD, we are building our lab culture as follows:
- Research-focused environment with minimal work outside of research
- Everyone, including undergraduate, works on their own project
- 1:1 meeting every week
- Enjoy interaction and discussion in the lab
- Encourage internal/external collaboration
- Use English in the lab
FAQ
- [Q] How many GPUs are availabe to me?
The Yonsei AI department has 600+ GPUs available (RTX3090, RTX4090, A6000, A100). We also have our own cluster with 16 L40S GPUs.
- [Q] Do you have robots?
We have a bimanual mobile robot from LG and will get a Franka Research 3 very soon. We also plan to order one Unitree H1 (or G1) humanoid robot.
- [Q] Which deep learning framework are you using?
Most previous projects were in PyTorch. But, we are mainly using JAX for current projects.
- [Q] Are there set working hours?
We have flexible working hours, but we encourage "core working hours (11am-5pm)" to facillitate interactions within the lab.
- [Q] Am I getting paid?
MS / PhD students and full-time interns will be paid a stipend that covers tuition, housing, and food.
- [Q] What is expected for undergrad interns?
Interns will work on their own research projects, individually or together with senior students.
- [Q] What group activities are there?
We have weekly group meetings, reading groups, and lab socials. We provide lunch for the group meetings and dinner for the lab socials!
- [Q] Why is English used in the lab?
You will read materials in English, write papers in English, and discuss and present your work in English. To help you become comfortable with these activities, we provide an English-friendly environment.
- [Q] Will there be any duties other than research?
We try to minimize your work outside of research (e.g. industrial projects, paperwork).
- [Q] Can I collaborate with researchers in the US?
Yes, we encourage external collaborations and internships at top research labs (e.g. Berkeley, Stanford, Google Deepmind). Look at our coauthors as a reference.
- [Q] I am not a Yonsei student. Can I apply?
Yes.
- [Q] How can I prepare for joining your lab?
If you are a Yonsei student, strongly recommend taking my courses, "Reinforcement Learning" and "Robot Learning". Alternatively, you can go through Berkeley's Deep Reinforcement Learning lectures. You may find our training materials for incoming students useful.