Published on June 4, 2025 10:19 AM GMT
TL;DR:
We're excited to announce the sixth iteration of ARENA (Alignment Research Engineer Accelerator), a 4-5 week ML bootcamp with a focus on AI safety! Our mission is to provide talented individuals with the ML engineering skills, community, and confidence to contribute directly to technical AI safety. ARENA will be running in-person from LISA from September 1st – October 3rd (the first week is an optional review of Neural Network Fundamentals).
Apply here to participate in ARENA before 23:59 on June 21st 2025 (anywhere on Earth).
Summary:
ARENA has been successfully run five times, with alumni going on to become MATS scholars and LASR participants; AI safety engineers at Apollo Research, METR, UK AISI, and even starting their own AI safety organisations!
This iteration will run from September 1st – October 3rd (the first week is an optional review of Neural Network Fundamentals) at the London Initiative for Safe AI (LISA) in Shoreditch, London. LISA houses AI safety organisations (e.g., Apollo Research, BlueDot Impact), several other AI safety researcher development programmes (e.g., LASR Labs, PIBBSS, Pivotal, Catalyze Impact), and many individual researchers (independent and externally affiliated). Being situated at LISA brings several benefits to participants, such as productive discussions about AI safety and different agendas, allowing participants to form a better picture of what working on AI safety can look like in practice, and offering chances for research collaborations post-ARENA.
The main goals of ARENA are to:
- Find high-quality participants;Upskill these talented participants in ML skills for AI safety work;Integrate participants with the existing AI safety community;Accelerate participants’ career transition into AI safety.
The programme's structure will remain the same as ARENA 5.0 (see below). For more information, see our website.
Also, note that we have a Slack group designed to support the independent study of the material (join link here).
Outline of Content:
The 4-5 week programme will be structured as follows:
Chapter 0: Neural Network Fundamentals
Before getting into more advanced topics, we first cover the basics of deep learning, including basic machine learning terminology, what neural networks are, and how to train them. We will also cover some subjects we expect to be useful going forward, e.g. using GPT-3 and 4 to streamline your learning, good coding practices, and version control.
Note: Participants can optionally skip this week of the programme and join us at the start of Chapter 1 if they’re unable to attend otherwise and if we’re confident that they are already comfortable with the material in this chapter. It is recommended that participants attend, even if they’re familiar with the fundamentals of deep learning.
Topics include:
- PyTorch basicsCNNs, Residual Neural NetworksOptimization (SGD, Adam, etc)BackpropagationHyperparameter search with Weights and BiasesGANs & VAEs
Chapter 1 - Transformers & Interpretability
In this chapter, you will learn all about transformers and build and train your own. You'll also study LLM interpretability, a field which has been advanced by Anthropic’s Transformer Circuits sequence, and work by Neel Nanda and the GDM Interpretability Team. This chapter will also branch into areas more accurately classed as "model internals" than interpretability, for example, work on steering vectors.
Topics include:
- GPT models (building your own GPT-2)Training and sampling from transformersTransformerLensIn-context Learning and Induction HeadsIndirect Object IdentificationSuperpositionSteering Vectors
Chapter 2 - Reinforcement Learning
In this chapter, you will learn about some of the fundamentals of RL and work with OpenAI’s Gym environment to run their own experiments.
Topics include:
- Fundamentals of RLVanilla Policy GradientProximal Policy GradientRLHF (& finetuning LLMs with RLHF)Gym & Gymnasium environments
Chapter 3 - Model Evaluation
In this chapter, you will learn how to evaluate models. We'll take you through the process of building a multiple-choice benchmark of your own and using this to evaluate current models through UK AISI's Inspect library. We'll then move on to study LM agents: how to build them and how to elicit behaviour from them.
Topics include:
- Constructing benchmarks for modelsUsing models to develop safety evaluationsBuilding pipelines to automate model evaluationBuilding and eliciting LM agents
Chapter 4 - Capstone Project
We will conclude this program with a Capstone Project, where participants will receive guidance and mentorship to undertake a 1-week research project building on materials taught in this course. This should draw on the skills and knowledge that participants have developed from previous weeks and our paper replication tutorials.
Here is some sample material from the course on how to replicate the Indirect Object Identification paper (from the chapter on Transformers & Mechanistic Interpretability). An example Capstone Project might be to apply this method to interpret other circuits, or to improve the method of path patching. You can see some capstone projects from previous ARENA participants here and here.
Call for Staff
ARENA has been successful because we had some of the best in the field TA-ing with us and consulting with us on curriculum design. If you have particular expertise in topics in our curriculum and want to apply to be a TA, use this form to apply. TAs will be well compensated for their time. Please contact info@arena.education with any more questions.
FAQs:
Q: Who is this programme suitable for?
A: There’s no single profile that we look for at ARENA; in recent iterations, successful applicants have come from diverse academic and professional backgrounds. We intend to keep it this way – this diversity makes our bootcamps a more enriching learning experience for all.
When assessing applications to our programme, we like to see:
- Applicants who genuinely care about AI safety and making the future development of AI go well;Applicants who are able to code well in Python, and have some knowledge of the maths needed for modern AI (linear algebra, calculus, probability);A solid understanding of how you might best contribute to technical AI safety, and how you expect ARENA to help you achieve your goals.
Since ARENA is an ML bootcamp, some level of technical skill in maths and coding will be required – more detail on this can be found in our FAQs. However, if our work resonates with you, we encourage you to apply.
Q: What will an average day in this programme look like?
At the start of the programme, most days will involve pair programming, working through structured exercises designed to cover all the essential material in a particular chapter. The purpose is to get you more familiar with the material in a hands-on way. There will also usually be a short selection of required readings designed to inform the coding exercises.
As we move through the course, some chapters will transition into more open-ended material. For example, in the Transformers and Mechanistic Interpretability chapter, after you complete the core exercises, you'll be able to choose from a large set of different exercises, covering topics as broad as model editing, superposition, circuit discovery, grokking, discovering latent knowledge, and more. In the last week, you'll choose a research paper related to the content we've covered so far & replicate its results (possibly even extend them!). There will still be TA supervision during these sections, but the goal is for you to develop your own research & implementation skills. Although we strongly encourage paper replication during this chapter, we would also be willing to support well-scoped projects if participants are excited about them.
Q: How many participants will there be?
We're expecting to accept around 30 participants in the in-person programme.
Q: Will there be prerequisite materials?
A: Yes, we will send you prerequisite reading & exercises covering material such as PyTorch, einops and some linear algebra (this will be in the form of a Colab notebook) a few weeks before the start of the programme.
Q: When is the application deadline?
A: The deadline for submitting applications is 23:59 on June 21st 2025, anywhere on Earth.
Q: What will the application process look like?
A: There will be three steps:
- Fill out the application form;Perform a coding assessment;Interview virtually with one of us, so we can find out more about your background and interests in this course.
Q: Can I join for some sections but not others?
A: Participants will be expected to attend the entire programme. The material is interconnected, so missing content would lead to a disjointed experience. We have limited space and, therefore, are more excited about offering spots to participants who can attend the entirety of the programme.
The exception to this is the first week, which participants can choose to opt in or out of based on their level of prior experience (although attendance is strongly recommended if possible).
Q: Will you pay stipends to participants?
A: We won't pay stipends to participants. However, we will be providing housing and travel assistance to in-person participants (see below).
Q: Which costs will you be covering for the in-person programme?
A: We will cover all reasonable travel expenses to and from London (which will vary depending on where the participant is from) and visa assistance, where needed. Accommodation, meals, and drinks and snacks will also all be included.
Q: I'm interested in trialling some of the material or recommending material to be added. Is there a way I can do this?
A: If either of these is the case, please feel free to reach out directly via an email to info@arena.education (alternatively, send us a LessWrong/EAForum message). We'd love to hear from you!
Links to Apply:
Here is the link to apply as a participant. You should spend no more than 90 minutes on it.
Here is the link to apply as a TA. You shouldn't spend longer than 30 minutes on it.
We look forward to receiving your application!
Discuss
