abbreviations before a period (Dr, Mr, e.g)
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. 20 of Qwen3-0.6B. Its strongest snippets are mixed: they peak on different tokens. 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 short abbreviations that are followed by a full stop, such as 'Dr', 'Mr', 'e.g' and 'Fig'. Turning it up did not produce a clear change in what the model wrote.
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.
-
it had absolutely nothing worry but fear itself” (pg. 6) c) “absolutely nothing to purchase and no money to purchase with it” (pg.⏎6) d) “nothing to see outside the boundaries of Maycomb County
-
said first and second control signals for input into said power management control circuit;⏎means connected to the output of said AC generator for developing a second error signal (e.g. e4) proportional to the difference in frequency between said AC generator and an
-
a uniform magnetic field as shown in Fig. (a).⏎(2) by rotating a magnetic field at a constant angular velocity within a stationary coil as shown in Fig. (a).⏎In either case, the generated voltage will be of sinusoidal
-
English. Others arrive with behaviour problems. Parents’ work schedules are often so inflexible that Federal Hill must cover basic logistics beyond school pick-up and drop-off. Ms Fernandez says she had to use her own car
-
two-fold; (1) to have students work with two contrasting design materials- one common and easily manipulated (e.g. wood) and one less common e.g plaster. (2) To cultivate hands-on material skills with the minimum of rapid
-
served in the cavalry in WWI. The flag flew over CDC for about 3 weeks, until the new one arrived. In 1995, Mr. Paine presented the flag to the museum. The 48-star,
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: [·····]*( · × ×°× ·°··°×°°°°°°°°°°°°°°°°°°°°
Most negative on
Where the activation went furthest below zero. Blue marks negative activation. Many neurons mean something different in each direction.
-
been given to you?⏎October 3rd, 2010, 12:50 PM⏎i started the homework⏎i dont know how to finish it here what i
-
-835-5322 (TTY) or at email@example.com.<|endoftext|>Table of Contents :⏎Top Suggestions Senior Kindergarten Writting Worksheet :⏎Senior Kindergarten Writting Worksheet In this worksheet your students will practice their handwriting
-
run using Run command. after selecting the path click on the next button.⏎Step 3: Set Run command⏎Now This is important to set run command.⏎Here you have to type the command name which you want to set.⏎Remember you can
-
value was changed to 20 in another(),so this prints 20. Thus final answer is 10 0 20. I hope i have been clear enough. ------------------ Come on in !! Drinks are on the house
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
- L19:184 strongly excites
- L18:454 strongly suppresses
- L18:580 excites
- L15:735 excites
- L18:232 excites
- L10:58 suppresses
- L16:704 suppresses
- L15:2567 excites
- L11:1820 suppresses
- L16:2959 suppresses
- L17:581 excites
- L14:160 suppresses
- L10:1050 excites end of document marker token
- L15:1287 suppresses
- L16:269 excites
- L10:2374 suppresses
- L17:119 suppresses
- L16:926 suppresses
- L19:268 excites
- L11:308 suppresses
- L15:2788 excites
- L18:804 suppresses
- L19:710 suppresses
- L9:1544 suppresses
Feeds later layers
- L22:120 excites
- L26:732 suppresses
- L25:643 excites
- L23:225 excites
- L25:409 suppresses
- L25:171 excites
- L23:291 excites
- L22:201 suppresses
- L21:1291 excites
- L24:190 excites
- L27:279 excites
- L27:243 excites
- L27:215 excites
- L26:146 excites
- L24:363 suppresses
- L26:89 excites
- L25:114 excites
- L25:522 excites
- L23:749 weakly excites
- L25:932 weakly excites
- L21:359 weakly excites
- L26:134 weakly excites
- L27:134 weakly excites
- L27:250 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
- ..
- l
- fie
- nn
- `
- .."
- s
- cls
- n
- amp
Suppresses
- onian
- ertime
- mPid
- ([[
- ��
- 鸰
- Seah
- 燧
- Pandora
- REW