Skip to content
BIASED
Open education · demo

Machine learning for people who want to understand what they’re doing.

Interactive courses with real intuition, math and practice. Two ways to read the same material. Immediate feedback. No fluff.

  1. 01Learn
  2. 02Practice
  3. 03Verify
  4. 04Recall
  5. 05Apply
  • Two learning modesPulse or Deep
  • Interactive practicewith instant feedback
  • Built for understandingnot memorization
  • Free for studentsacademic access

One source, two ways to read it

Pulse and Deep are not two different texts. They render the same source: the same equations, the same code, the same questions. Switch at any time — your answers and progress stay.

Pulse

Short blocks with questions

A 1–3 minute block → 1–3 checks → feedback → next block. For the commute, a break, or a topic that still feels hard.

Deep

Full topic, then questions

The whole topic as one coherent text with derivations, examples and code. Questions collected at the end. For reading at a desk.

The learning loop

Reading is not understanding. So a topic ends with a check, not with “read”.

  1. 01Live

    Learn

    Theory in short blocks or in one piece

  2. 02Live

    Practice

    Short tasks, 30 seconds to 7 minutes

  3. 03Live

    Verify

    Deterministic grading on the server

  4. 04Planned

    Recall

    Delayed review a few days later

  5. 05Planned

    Apply

    Projects and Python tasks

Courses

Structured, interactive and practice-driven.

All courses

Generative Models

Preview

VAEs, GANs, flows and diffusion — derived, not hand-waved.

1 of 15 topics available

Deep Learning

Coming soon

From neural networks to modern architectures.

—

Mathematics for ML

Coming soon

Linear algebra, probability and optimization, with intuition and practice.

—

ML in Practice

Coming soon

Build, experiment and understand real-world systems.

—

Prior — a companion that hints but never solves

A Bayesian cat lives next to the exercises. A hint is always one click away and comes in steps: an idea, then a guiding question, then an intermediate step. Prior never gives away the full answer — and honestly marks that you used a hint.

Prior

a Bayesian cat

  1. 1Nudge
  2. 2Guiding question
  3. 3Scaffold
  4. ✕The full answer

Theory is open. You pay for convenient practice.

Public notes stay free. The subscription covers the interactive part: checks, review and projects.

See pricing