variables and symbols in formulas
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. 14 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 fires on the letters and symbols in maths and science formulas, like the L in 'LdB', the P in 'P(t)', or the E in 'E=hf'. Turning it up made the model write equations with equals signs and numbers, which fits the same theme.
Written by Claude from this neuron's activations and steering outputs. Confidence: high. 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.
-
value for pixels in image I be MAX.⏎Express the PSNR in decibels. From Step 1, we have the decibel value LdB as LdB = 10 log10 (P1/P0). Now let
-
. The phase-difference can, therefore, be defined as ΔΦ=ΦV-ΦI.⏎We define the instantaneous power similarly to the DC power by P(t)=V(t)×I(t). When using
-
existing asset base to satisfy its creditors, owners, and other providers of capital, or they will invest elsewhere.⏎The following is the WACC calculation formula:⏎WACC = E/V × Re + D/V × Rd × (1 - Tc
-
sphere is at distance R from the center. What is the centripetal acceleration of the same sphere if both distance R and speed s are doubled?⏎B) a / 4⏎C) a / 2⏎D) 2 a
-
waves ( waves property of particle) and its speed ( frequency) is : c>1. We call such particle - ? electron? and its energy is: E=h*f.⏎In this way I can understand the reality of nature. ==.
-
,- 1) are orthogonal vectors in R 4 . Write the vector X = (1 , 1 , 1 , 2) in the form X = a U +<|endoftext|>Hepatitis B vaccine⏎The hepatitis B vaccine (
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: what the world, the idea of the world, the world, the culture, the meaning of the world, the world, the culture, the世界, the世界,的世界,的世界,的世界,的世界,的世界
Most negative on
Where the activation went furthest below zero. Blue marks negative activation. Many neurons mean something different in each direction.
-
10th - 12th English Language Arts⏎Lesson Plan: The State of the Union⏎Learners explore the state of the union. In this U.S. government lesson, students analyze the 1998 State of
-
is no sign of toxicity from the composite.<|endoftext|>Meters to millimeters (meter to mm converter) returns numbers of mm giving meters m value. Converting 1 meter to millimeter as 0.001. See also mill
-
the heat source⏎- lateral current at bottom of water TOWARD the heat source.⏎A. Setup (download Quicktime; 1.6 Mb)⏎B. Convection in action (download .mp4; 22 Mb)⏎
-
collections to collaborate.⏎Organize your curriculum with collections. Easy!⏎Have time to be more creative & energetic with your students!⏎Students explore the Orange Revolution. In this Orange Revolution lesson, students examine the climate of the Ukrainian protest and investigate the role
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
- L9:31 excites
- L7:1089 weakly excites
- L3:48 weakly excites
- L7:323 weakly excites
- L8:687 weakly excites
- L13:532 weakly suppresses
- L8:137 weakly excites
- L11:2558 weakly suppresses
- L11:894 weakly suppresses
- L10:657 weakly suppresses
- L9:280 weakly suppresses
- L13:1262 weakly excites
- L8:1828 weakly excites
- L11:164 weakly excites
- L13:347 weakly suppresses
- L8:2184 weakly suppresses
- L10:588 weakly excites
- L1:1960 weakly excites
- L10:2686 weakly suppresses
- L7:798 weakly excites
- L4:24 weakly suppresses
- L13:1457 weakly suppresses
- L11:1631 weakly excites
- L12:1263 weakly suppresses
Feeds later layers
- L15:2416 excites
- L16:943 excites
- L26:606 excites
- L18:1145 excites
- L24:83 suppresses the digit 1 starting a year
- L27:390 excites
- L16:1581 excites
- L16:2165 excites
- L19:2749 excites
- L25:252 excites
- L20:772 weakly excites
- L16:1947 weakly suppresses
- L19:1486 weakly excites
- L15:2090 weakly excites
- L24:185 weakly excites
- L18:1061 weakly suppresses
- L17:1231 weakly excites
- L16:1280 weakly excites
- L16:1680 weakly excites
- L19:1732 weakly excites
- L15:1830 weakly suppresses
- L19:1093 weakly excites
- L27:2825 weakly suppresses
- L19:2219 weakly suppresses
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
- =
- =
- 他们在
- =”
- ={
- emain
- =(
- ="
- =(
- =[
Suppresses
- ographically
- crets
- フト
- 个百分
- acher
- odash
- ibrator
- htags
- '}⏎
- aidu