Observed arrival · 2026-10-30
Inside Qwen3, one neuron at a time
Neuron Explorer presents a map of all 86,016 neurons in the open-weight Qwen3-0.6B language model, with tools for exploring their responses and steering their effects.
○Open
⊠Login
$Paid
†Ads
✦Pretty
●Pro
◎Niche
◉Human
⚑Risk
ƒJS
- For
- People exploring language-model internals and interpretability
- Worth noticing
- The map is built from precomputed activations; the page says nothing runs on its server.
Field notes
Qwen3-0.6B is described as having 28 layers, with 3,072 neurons in each; the explorer says its map is built from real activations and precomputed data, so inference does not run on the server. The interface offers Map, Tour, Play, Learn, and neuron search, alongside Simple and Technical views; the extracted page shows the map still loading.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
○OpenPublic substance visible
✦PrettyNotable craft visible
●ProPolished or operationally mature
◎NicheUnusually specific use
One card from the complete issue