Collective Intelligence
2025/26/2
Spring 2026 edition of the Collective Intelligence practice series, continuing with my PhD colleague Zoltán Barta. The practice notebooks are the same battle-tested set from the previous semester. What’s new is a redesigned Assignment 1 with an open-ended NetLogo prompt, a tighter Assignment 2 with four hand-picked MARL environments, and a new practice session on handling heterogeneous teams in MARL.
This time we’re also running a small special prize for the most creative visualisation in Assignment 1, purely for fun and to keep both halves of the brain working.
Please note that the materials may contain small mistakes, typos, or even implementation bugs. I would appreciate any notifications about these issues sent to my email.
Practice Content
All practice materials live on Google Colab, accessible via cloud links with no local setup required.
For the first assignment defence, every student submits their complete implementation in .nlogo format with an accompanying short presentation, delivered in class. For the second assignment, all students collaborate on a GitHub project hosted on the CI GitHub Organization. Same defence rules apply: working code + a short presentation.
Course introduction. Logistics, assignments, and the Assignment 2 topic list.
Introduction to NetLogo and the Logo programming language. Assignment 1 announcement.
Introduction to NetLogo Models Library and simulations.
Designing NetLogo models from scratch. Design principles.
Assignment 1 defence and presentations.
Introduction to Games and Models for Multi-Agent Interactions. Assignment 2 announcement.
Introduction to TorchRL and the Centralized Training Decentralized Execution (CTDE) Framework.
Handling Heterogeneous Teams in Multi-Agent Reinforcement Learning.
Assignment 2 defence and presentations.
Assignments
The course is structured around two core assignments. The first is an open-ended NetLogo project focused on ideation and the development of an agent-based model. The second is a small enterprise-level group project in which students collaboratively design and implement a solution for one of four hand-picked multi-agent reinforcement learning environments.
This semester’s bonus: the instructors pick a winner for the most creative visualisation in Assignment 1, with a small surprise prize handed out at the Assignment 1 defence.
Designing a custom ABM in NetLogo based on constraints. Creative-visualisation prize awarded at defence.
Group MARL project on one of four environments. Built with TorchRL and submitted via the CI GitHub Organization.
Exams
The grade is determined by the results of two assignments, which together create a composite grade. Additionally, to receive this grade, students must pass an exam based on the lecture material. These exams are administered by the Course Owner, László Gulyás. Here are this semester’s mock exams:
Course Syllabus
Schedule
- Lecture: Wednesdays, 16:00 - 18:00, South Building, Room 0-817 (Hungarian: Déli Tömb 0-817)
- Practice: Wednesdays, 14:00 - 16:00, South Building, Room 00-807 (Hungarian: Déli Tömb 00-807)
Description
This course is designed to provide students with an in-depth exploration of topics related to Collective Intelligence. Students will learn about agents, agent-based systems, and Multi-Agent Systems, along with their relevance to real-life dynamics and systems.
Participants will gain a comprehensive understanding of the Logo programming language and the NetLogo software, which will enable them to create simulations using agent-based modeling. Additionally, the course covers Game Theory, Algorithmic Game Theory, and foundational algorithms such as Minimax Q-Learning, Nash Q-Learning, Fictitious Play, and Neural Replicator Dynamics.
Moreover, students will delve into Multi-Agent Reinforcement Learning using the TorchRL framework. They will create simple systems utilizing the IPPO and IDQN algorithms to tackle toy and benchmark problems. Throughout the course, students will learn how to tune reward systems, debug multi-agent reinforcement learning (MARL) systems, perform environment transformations, implement feature extractor networks, and (new this semester) handle heterogeneous teams of agents with mixed roles.
Grading
Your final grade is calculated using the formula below:
Final Score = Assignment 1 (30 points) + Assignment 2 (70 points)
| Final Score Range | Grade |
|---|---|
| > 85 | 5 |
| 75 - 85 | 4 |
| 65 - 74 | 3 |
| 40 - 64 | 2 |
| < 40 | F |
- Pass required on a final written exam from the lecture material (Pass/Fail Exam)
Prerequisites
- Python (moderate level)
- Linear Algebra (moderate level)
- Reinforcement Learning Concepts (advantageous but not required)
- Deep Learning Concepts (moderate level)
Tools and Frameworks
- Programming Language: Python, Logo
- Frameworks: PyTorch, TorchRL
- Libraries: NumPy, Gymnasium, TorchRL, Hydra
- Additional Tools: Google Colab, NetLogo, Docker, Git
Learning Objectives
- Understand the basics of Multi-Agent Systems
- Understand the Logo programming language and the utilization of the NetLogo software
- Understand how Game Theory is connected to Collective Intelligence
- Be able to implement a MARL solution for a toy problem
Recommended Reading
- Wooldridge, M. (2009). An Introduction to Multi-Agent Systems. Wiley.
- Axelrod, R. (1984). The Evolution of Cooperation. Basic Books.
- Osborne, M. J. & Rubinstein, A. (1994). A Course in Game Theory. MIT Press.
- Russell, S. & Norvig, P. (2020). Artificial Intelligence: A Modern Approach.
- Sutton, R. S., & Barto, A. G. (2018). Reinforcement Learning: An Introduction.
- NetLogo Documentation: Center for Connected Learning and Computer-Based Modeling.
- Nisan, N., Roughgarden, T., Tardos, É., & Vazirani, V. V. (2007). Algorithmic Game Theory. Cambridge University Press.
back to teaching