NEXT BRAIN Computational Neuroscience & NeuroAI

Next Brain · Explore

Explore neural systems

Change a parameter, watch neurons fire and follow an image into neural signals. Explore interactive simulations and recorded activity, with links to the models and evidence behind each experience.

LiveBrowser laboratory

01 / Neural Lab

Change a parameter.
Watch neurons fire.

Explore established neuron models, write custom equations and run bounded simulations entirely in the browser.

  • Adaptive LIF, Izhikevich, Hodgkin–Huxley and more
  • Equation Lab and Brian2 model import
  • Local WASM reference and WebGPU execution
ExecutionOn device
DataStays local
Open Neural Lab
LiveConnectome experiment

02 / FlyLab · Digit Arcade

One connectome.
Ten possible answers.

Test digit recognition built around a fixed FlyWire-derived network, with a playful interface and explicit scientific boundaries.

  • 15.1M weighted connectome edges
  • MNIST challenge and freehand input
  • On-device or connected server execution
ExecutionLocal + remote
Interface3D arena
Open Digit Arcade
Recorded activityVision experiment

03 / Vision Lab

Follow an image.
Inspect its spikes.

Explore visual encoding and STDP digit learning through real Brian2 recordings, with pause, step and individual neuron inspection.

  • 768 pixel neurons and measured voltage traces
  • MNIST responses and learned weight checkpoints
  • Short teaching experiments with source provenance
Pixel → spike Spike → digit
ExecutionRecorded replay
DataBrian2 simulation
Open Vision Lab

Built as real applications

One platform identity.
Purpose-built interfaces.

A lab is not an embedded demo. It gets the runtime, controls and assets its scientific question requires, while the platform supplies the path back to reproducible work.

Its own runtime

Workers, WASM, WebGPU and specialist assets load on the Lab's route—not in the platform shell.

Explicit boundaries

Every experience explains what is computed, what is visualised and what the result does not prove.

Connected provenance

Models, parameters and results can move into the shared version graph without flattening the experience.

Continue the work

A Lab is an entry point—not an island.

Find the modelInspect lineage and supported backends.Build in the workspaceFork a model and change the experiment.Review the runKeep parameters, resources and results together.Read the evidenceSee what is known and what remains open.
From browser to clusterThe same scientific identity can travel beyond the Lab.
See how the platform works