AboutWhat I doProjectsLabContact
ManifestoHow machines learnBlogOpen SourceThe numbersWhat is out thereThe pathThe map of everythingMonthly challengePlaygroundRunnable notebookUsesNowResumeShall we talk?
LabRuns entirely in your browser. Nothing is sent to any server.

The agent that learns to be afraid

An agent crosses a maze on rewards and punishments, with its value table painted underneath: fear of the traps shows up before the path to the goal, because punishment arrives in a single update while reward has to propagate. Edit the walls and watch it adapt.

Reward and punishment, told without jargon

More experiments:

Shall we connect?

Let's talk: feedback, collaboration or an opportunity.

You can write to me about a project, a technical question, to give me feedback, or about an internship or a first junior role. I always reply.

Write to me

hola@jmwebsoluciones.com