MLx Representation Learning & Generative AI

Oxford Machine Learning School

Discover the latest advancements in representation learning, and stay informed about the continuous evolution of generative AI and emerging foundation/frontier models.

What you'll be able to do

  1. 01 Can justify geometric and multimodal representation choices using data symmetries, equivariance and relationships between modalities.
  2. 02 Can compare flow matching, diffusion and other generative approaches, explaining how probability paths and optimal transport affect training and sampling.
  3. 03 Can select Bayesian data-gathering objectives and interpret causal assumptions to support decisions under uncertainty and distribution shift.
  4. 04 Can analyse multi-agent coordination problems and recommend learning, communication and human-alignment strategies, identifying scalability and safety trade-offs.
  5. 05 Can critically evaluate foundation models using benchmarks, cost-aware comparisons, probes and activation patching, distinguishing predictive signals from causal evidence.
  6. 06 Can assess world-model and AI-assisted discovery workflows, including mathematical reasoning, identifying where verification and formal proof are needed.

12 modules, one certification exam

Your learning plan opens in your workspace, with the unit you are on always marked up next.

Certificate Oxford Machine Learning School

MLx Representation Learning & Generative AI

12 modules 13 units 18h 12m

Continue Learning Up next: Open Endedness, World Models and The Automation of Innovation
  1. Module 01

    Open Endedness, World Models and The Automation of Innovation

    This module examines how AI is moving beyond narrow automation toward open-ended systems that can generate novel artifacts, learn in rich environments, and support scientific discovery. It traces the shift from early automation ideas to foundation models, explains why game-like simulation success does not transfer directly to the real world, and introduces world models as controllable simulators for embodied learning. It also explores how scaling, curriculum learning, and latent actions can expand capability, and how foundation models may help automate innovation through iterative generate-test-refine loops.

    1 Unit 1H 44M
    1. U01 Open Endedness, World Models and The Automation of Innovation Up next
  2. Module 02

    Benchmarking

    1 Unit 1H 28M
  3. Module 03

    Multi-Robot and Multi-Agent Learning

    2 Units 1H 43M

Inside the MLx Representation Learning & Generative AI learning plan

12 modules · 13 units· ~18h of structured prep.

  1. 01

    Open Endedness, World Models and The Automation of Innovation

    1 unit · ~2h
    1. 1. Open Endedness, World Models and The Automation of Innovation
  2. 02

    Benchmarking

    1 unit · ~1h
    1. 1. Benchmarking
  3. 03

    Multi-Robot and Multi-Agent Learning

    2 units · ~2h
    1. 1. Part 1: Multi-Robot and Multi-Agent Learning
    2. 2. Part 2: Multi-Robot and Multi-Agent Learning
  4. 04

    Intelligent Data Gathering

    1 unit · ~2h
    1. 1. Intelligent Data Gathering
  5. 05

    On Causal Learning: Discovery and Extrapolation

    1 unit · ~2h
    1. 1. On Causal Learning: Discovery and Extrapolation
  6. 06

    Multimodal AI

    1 unit · ~2h
    1. 1. Multimodal AI
  7. 07

    Human-Centric Cooperative AI: Evaluating, Aligning & Governing Multi-Agent Systems

    1 unit · ~1h
    1. 1. Human-Centric Cooperative AI: Evaluating, Aligning & Governing Multi-Agent Systems
  8. 08

    Continuous-Time Generative Modeling: Matching and Optimal Transport

    1 unit · ~1h
    1. 1. Continuous-Time Generative Modeling: Matching and Optimal Transport
  9. 09

    The Rise and Ramifications of Artifical Mathematical Intelligence

    1 unit · ~1h
    1. 1. The Rise and Ramifications of Artifical Mathematical Intelligence
  10. 10

    Geometric Deep Learning

    1 unit · ~1h
    1. 1. Geometric Deep Learning
  11. 11

    Why Formalize Mathematics?

    1 unit · ~1h
    1. 1. Why Formalize Mathematics?
  12. 12

    Probing Neural Networks

    1 unit · ~1h
    1. 1. Probing Neural Networks

How the content is written

Worked examples
None
Technical jargon
None
Maths & formulas
None
Code
None
Diagrams
None

How you certify

Pass the certification exam to earn your certificate.

Questions
100
Time limit
120min
To pass
70%
  • Single select (Radio): 4 options
  • Multiple select (Checkbox): 4 options
  • True / False
Issued by
Oxford Machine Learning School OxML
Certificate type
Certificate of Completion
CPD accredited
CPD Certified Certification number 77710
Programme dates
15 – 18 July 2026

The certificate you earn

Issued in your name, with a verification link anyone can check. Download it as a PDF or add it to LinkedIn from your award page.

Start your MLx Representation Learning & Generative AI certificate.

Start working toward MLx Representation Learning & Generative AI for free today.