Brain-like computation and intelligence

NX-414

This file is part of the content downloaded from Brain-like computation and intelligence.

Welcome to NX-414!

Recent advances in machine learning have contributed to the emergence of powerful models for how humans and other animals reason and behave.

In this course we will compare and contrast how such brain models as well as brains create intelligent behavior. Specifically, we will cover the following topics:

Class	Date	Topic	Exercise session
1	19/02/2025	Introduction & neural code
2	26/02/2025	Normative models
3	05/03/2025	Bayes and Brain-like circuits
4	12/03/2025	Task-driven models (Path integration)
5	19/03/2025	Task-driven models (Vison)
6	26/03/2025	Task-driven (Unsupervised, Audition, metamers, optimal stimuli)	Project
7	02/04/2025	Task-driven (Proprioception)	Project
8	09/04/2025	Language modeling in the brain I	Quiz (presence required)
9	16/04/2025	Language modeling in the brain II	Project
10	23/04/2025	EPFL Easter break 🥚🌸
11	30/04/2025	Motor Control	Project
12	07/05/2025	Language modeling in the brain III (language in the service of cognition)
13	14/05/2025	Brain-inspired reinforcement learning
14	21/05/2025	Skill learning	Quiz (presence required)
15	28/05/2025	Review	Review session

Teaching team

Alexander Mathis – alexander.mathis@epfl.ch

Martin Schrimpf – martin.schrimpf@epfl.ch

Teaching assistants:

Michael Hauri (Research assistant)

Abdulkadir Gokce (PhD student)

Hossein Mirzaei Sadeghlou (PhD student)

Merkourios Simos (PhD student)

Learning Prerequisites

CS-433 (strongly recommended)
Programming in Python, good mathematics and machine learning background

Learning Outcomes

By the end of the course, the student must be able to:

Formulate models of brain function
Hypothesize potential mechanisms that give rise to behavior
Design models of brain functions
Characterize current models of brain function

Transversal skills

Set objectives and design an action plan to reach those objectives.
Demonstrate the capacity for critical thinking
Write a scientific or technical report.
Summarize an article or a technical report.

Teaching methods

Lectures and exercises to discuss and work on problem sets (both numerical and analytical).

Expected student activities

Attend lectures and take notes during lectures, participate in quizzes and read scientific articles. Participate in the coding project, solve the problem sets and take the final exam.

Some weeks include problem sets with pen-and-paper exercises. Students are encouraged to go through these exercises in advance. Exercise sessions will serve to delve deeper into the material, address questions, and present the solutions.

Assessment methods

The final mark is a combination of three evaluations: problem sets/project (25%), quizzes (25%), final exam (50%).

Brain-like computation and intelligence

February 19th