"the" introducing a noun phrase after a name or preposition
A neuronOne of thousands of tiny dials inside each layer. It reads the text so far and produces a number, its activation, that nudges what the model writes next. in layerOne step in the stack the text passes through. Qwen3-0.6B has 28 layers, each with 3,072 neurons. 6 of Qwen3-0.6B. Its strongest snippets are very consistent: they mostly peak on the same token. It is featuredA neuron picked for the full treatment: steering outputs and a written explanation. About 3,000 of the 86,000 neurons are featured., so it has steering outputs.
What this neuron does
This neuron lights up on the word "the" when it comes right after a name or a phrase like "with" or "in", as in "Rosey the Riveter" or "the man in the black hat". When it was turned up, the writing drifted into odd, repetitive loops.
Written by Claude from this neuron's activations and steering outputs. Confidence: medium. Think it is wrong? Flag it (opens a prefilled issue on GitHub).
Fires most on
The text windows where this neuron's activationThe number a neuron produces for a token. Large positive means it fired hard; negative means it pushed the other way. was highest. Amber marks positive activation; the outlined tokenA piece of text the model reads at once: often a word, sometimes part of a word or a punctuation mark. is the peak. Hover a token for its strength.
-
1000-piece puzzle of World War II propaganda posters tells the story from the Home Front. Who wouldn't recognize Rosey the Riveter, the woman, sporting a red bandana, her bicep
-
caused many streetcars to be stalled behind the first, now horseless, car. Bergh then stopped a car going in another direction.⏎The “man in the black hat” had been trying for many months to get the streetcar companies to
-
''s most intelligent animals. See a great silverback gorilla, the monkey with the biggest nose, how gibbonsswing through the trees, multicolored m
-
race? What would it be like? How would it feel?⏎is one of the fascinating scenarios Philip K. Dick exposes us to in his classic The Man in the High Castle. This 1962 alternative history portrays a world in which
-
, I can't change a⏎> component of somethingNew without creating y. The point is there's⏎> nothing about x that changed,⏎The thing with the changed component is not x anymore.⏎> and there
-
ball boy watches over them.⏎Do kickers get tackled?⏎Yes, the kicker and punter are expected to try to make a tackle, if the person with the ball gets that far. … You can tackle the kicker / punter if you
SteeringAdding a push along a neuron while the model writes, to see how its output changes. Turning it up or down.
Push this neuron up or down while the model writes, and watch what changes. The outputs were generated ahead of time, so nothing runs in your browser.
Here is something I have been thinking about lately: "If you are given the choice of your own hands to choose your own heart's desire...Youthful Medicine." This is a common statement, but the question is, does it really work or not. The intent of your heart,
Most negative on
Where the activation went furthest below zero. Blue marks negative activation. Many neurons mean something different in each direction.
-
construction for the new Heilwood Company store had begun. It would be a two-story, wooden frame building and included not only a general store, but also the town post office. The store manager’s house was constructed directly below the store (
-
to a garden, a theme to which all other parts relate. This may be a walk or a summer-house or a sun-dial or a garden bed or the residence itself, or a brook falling down the sward between trees and bushes
-
-years wide⏎Isaac Newton’s universe was a cozy, tidy place. Gathered around the sun were six planets, a handful of moons and the occasional comet, all moving against a backdrop of stationary stars. Newton provided us with
-
nature⏎a person who rules with total and unlimited control⏎a narrow waterway connecting two larger bodies of water⏎a state made up of an independent city and the nearby countryside⏎a government by a few⏎a government by the people⏎a
Connections
Show on mapWhich neuronsOne of thousands of tiny dials inside each layer. It reads the text so far and produces a number, its activation, that nudges what the model writes next. in earlier layersOne step in the stack the text passes through. Qwen3-0.6B has 28 layers, each with 3,072 neurons. feed this one, and which later ones it feeds. Amber links excite, blue links suppress; longer bars are stronger.
Fed by earlier layers
- L4:52 excites
- L4:73 suppresses
- L5:125 excites "of" after all three, both, ten
- L5:1747 weakly suppresses
- L4:19 weakly suppresses
- L4:731 weakly excites
- L4:49 weakly suppresses
- L4:683 weakly suppresses
- L3:1456 weakly suppresses
- L5:225 weakly excites
- L2:2268 weakly excites
- L4:245 weakly excites
- L3:1447 weakly suppresses
- L1:2419 weakly excites
- L0:2802 weakly suppresses
- L4:962 weakly excites
- L2:721 weakly excites
- L3:1784 weakly excites
- L4:130 weakly excites
- L0:912 weakly excites
- L0:2718 weakly excites
- L3:291 weakly excites
- L5:561 weakly excites
- L5:9 weakly suppresses
Feeds later layers
- L9:656 suppresses
- L9:24 suppresses "the" before a named organisation or title
- L8:466 excites 'a' or 'an' before a noun phrase
- L11:875 suppresses
- L20:901 excites
- L16:2062 suppresses "other" and contrast words
- L18:314 suppresses
- L17:188 excites 'a' before a noun phrase
- L12:1990 suppresses
- L17:1542 excites
- L12:435 suppresses
- L18:2305 weakly excites
- L16:1309 weakly excites
- L18:884 weakly excites
- L17:30 weakly suppresses
- L16:2248 weakly excites
- L14:2374 weakly excites
- L18:635 weakly excites
- L20:1123 weakly suppresses
- L12:1364 weakly suppresses
- L20:2682 weakly excites
- L9:59 weakly excites "a" meaning per (a day, a year)
- L19:1475 weakly excites
- L20:1907 weakly excites
Activation histogramA chart of how often a neuron takes each activation value.
How often this neuron fires at each strength. Almost always it does close to nothing (the tall bars in the middle); the snippets above are the rare moments it fired hard, amber for positive and blue for negative.
Logit lensReading which output tokens a neuron pushes toward or away from, straight off its output weights, before the later layers have their say.
The words this neuron nudges the model toward (amber) or away from (blue) when it fires. A hint, not a label: many neurons work through later layersOne step in the stack the text passes through. Qwen3-0.6B has 28 layers, each with 3,072 neurons. instead.
Promotes
- alto
- altet
- _strip
- éis
- 企业提供
- 蹼
- achten
- cripts
- raid
- gett
Suppresses
- 家乡
- std
- able
- ::
- 我家
- 我省
- 一名
- 堪
- aspects
- 臺南市