document end markers and disc brakes
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. 7 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 is mixed. It fires most on the hidden 'end of document' marker that web pages use between unrelated texts, and also on the phrase 'disc brakes'. Its negative side is weak and hard to read. Turning it up or down did not produce a clear theme in the model's writing.
Written by Claude from this neuron's activations and steering outputs. Confidence: low. 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.
-
Nov 23, 2021, 1 PM⏎World StampsNov 22, 2021, 7 PM<|endoftext|>At SACRE, the Fluent CFD code has been used in the past
-
please your palate for centuries to come.⏎KONISHI BREWING CO., LTD.⏎3-5-8, Chuo, Itami city,<|endoftext|>Materials
-
, Light and Space of the Medieval Cathedrals of England and France" in 1973 (Aperture).⏎|[Checklist]||Click<|endoftext|>The unmanned Jules Verne ATV cargo ship breaks up in a spectacular display⏎
-
as are the essence of Mexican food⏎- What Is the Difference Between Fruits and Vegetables?⏎Try Our Delicious Dishes!⏎Contact us today by telephone (<|endoftext|>One food which dates back in prehistoric times is the date. Here is truly
-
782-7937)⏎For more information on activities ongoing within EPA Region 7 (Iowa, Kansas, Missouri and Nebraska), contact:<|endoftext|>According to the analysis of the cutting process oftooth surface, based on the
-
," he said.<|endoftext|>Disc Brake Rotor Specsby Craig Woodman⏎Four-wheel disc brakes have become more common in vehicles because of their increased stopping power. Disc brakes are a reliable system, but still need routine maintenance. In addition to replacing
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: I have never been aware that the car is equipped with a certain model of gear shift, the car is equipped with a certain model of foot, or the car is equipped with a certain model of wheel to the left or right position for the
Most negative on
Where the activation went furthest below zero. Blue marks negative activation. Many neurons mean something different in each direction.
-
a review, the scientific knowledge about fitness for transport is reviewed against the background of the relevant European legislation.⏎Injury and unnecessary suffering⏎The European regulation (Council Regulation 1/2005 and to a lesser extent 853
-
by By Dr. Carolyn K. Stratz⏎Q: My kids get their vaccinations - what about me?⏎A: Back-to-school time
-
- Quiz Questions & Answers⏎- Which Italian inventor succeeded in transmitting wireless waves over long distances for the first time?⏎-
-
long does it take for vision therapy to work?⏎Most children begin actually noticing a difference within six to eight weeks, about midway through the typical therapy program. Q: Will my child’s vision problem come back later in life? A: Generally speaking
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
- L6:1825 suppresses
- L3:105 weakly suppresses
- L5:1498 weakly suppresses
- L5:78 weakly suppresses
- L5:458 weakly excites
- L5:741 weakly excites
- L4:1542 weakly suppresses
- L3:894 weakly suppresses
- L3:3049 weakly excites
- L4:2510 weakly suppresses
- L2:570 weakly suppresses
- L5:1391 weakly suppresses
- L2:1965 weakly suppresses
- L6:1083 weakly suppresses
- L1:2726 weakly suppresses
- L2:2602 weakly suppresses
- L4:1051 weakly suppresses
- L6:892 weakly suppresses
- L1:1728 weakly suppresses
- L5:198 weakly excites unclear: weak "and" in lists
- L4:1691 weakly excites
- L5:1636 weakly suppresses
- L3:237 weakly excites
- L3:898 weakly suppresses
Feeds later layers
- L8:950 excites
- L20:1413 suppresses
- L9:498 weakly excites
- L9:2341 weakly suppresses
- L19:846 weakly excites
- L10:291 weakly excites
- L9:720 weakly suppresses
- L18:1990 weakly suppresses
- L27:1203 weakly excites
- L9:33 weakly suppresses
- L12:1840 weakly excites
- L22:312 weakly suppresses
- L17:2136 weakly suppresses
- L9:128 weakly excites
- L17:1881 weakly suppresses
- L14:1839 weakly suppresses
- L17:1629 weakly suppresses
- L21:429 weakly suppresses
- L14:2556 weakly suppresses
- L24:2543 weakly suppresses
- L18:2610 weakly suppresses
- L14:650 weakly excites
- L15:552 weakly excites
- L12:1804 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
- ())[
- grav
- ��이
- )[-
- 랩
- знания
- ()};⏎
- ());⏎⏎⏎
- ();⏎⏎⏎
- ");⏎⏎⏎
Suppresses
- بارك
- 嚏
- roys
- moot
- aż
- /history
- oyal
- 茬
- licensors
- :UIAlert