EXPERIMENTS / PLAYGROUND

Machine learning for sensing and robotics.

Interactive studies in learning systems, sensing, simulation, and the software that connects them to the physical world.

The live studies expose their controls and output so you can inspect what changes.

Neural Field

Watch a browser-trained field learn a surface in real time.

How it learns
A SMALL MODEL. A LIVING SCULPTURE.

Train a neural field.

A neural network learns to turn two surface coordinates into a point in three dimensions. Each particle follows its predictions.

Watch the shape emerge from its first predictions. Switch forms to give it a new shape to learn. Amber marks the places where it still differs from the mathematical target.

824243

Trained here in your browser. No API, uploaded data, or pretrained model. Cursor movement bends the view; it does not train the network.

INITIALIZING MODEL
—FITTING ERROR
Sculpted flowMove your cursor. Disturb the field.
Explore the Neural Field study

ALL EXPERIMENTS

Learning systems, sensing, robotics, and interfaces.

All experiments

LIVE TACTILE STUDY

Read the research journey

INTERACTIVE INFERENCE

Predict fingertip contact and force.

This browser experiment models a synthetic magnetic fingertip. Move the contact, change its force, then replay its small learning model.

TRY THE EXPERIMENT

Train a model to estimate contact and force.

Teach a small network to read a synthetic magnetic fingertip.

Magnetic tactile sensing: contact, magnetic displacement, and measurementA cutaway diagram of a compliant sensing surface. A probe indents the elastomer, shifting embedded magnets relative to a fixed sensor board. Magnetic measurements feed the contact-estimation model. This explains the principle and is not a measured research result.APPLIED FORCEELASTOMERMAGNETSSENSORSINFERENCE
System diagram: contact deforms the sensing surface; a learned model estimates position and force.