Neuron Explorer

Qwen3-0.6B

colon before a list or answer

A neuron in layer 16 of Qwen3-0.6B. Its strongest snippets are very consistent: they mostly peak on the same token. It is featured, so it has steering outputs.

What this neuron does

This neuron lights up on a colon at the end of a line, right before a list, a set of reasons, or a question and answer begins. Turning it up mostly produced rambling or off-topic text, so the link to the trigger is only a guess.

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 activation was highest. Amber marks positive activation; the outlined token is the peak. Hover a token for its strength.

  1. #1 peak

    . Take a look below at some of the most common questions that we hear from educators and the solution to solve them:⏎"Where do I start?"⏎You are an educator who either purchased your own classroom

  2. #2 peak

    are several key reasons why copper stills are highly valued:⏎Excellent Thermal Conductivity of Copper:⏎Copper possesses outstanding thermal conductivity, surpassing

  3. #3 peak

    that is likely caused by many factors. Some of these include:⏎The Brain May Be Cause of Nausea⏎Copyright © Nucleus Medical

  4. #4 peak

    know, here's some things to check:⏎Its going to vary by manufacturer, window type, and age, but some basic

  5. #5 peak

    including:⏎· Species distribution and abundances⏎· Species movements and life stages⏎· E

  6. #6 peak

    deconstruct, and comprehend music. This helpful guide will give you a great grasp of:⏎- Note value and counting notes⏎- Treble and bass clefs⏎-

Steering

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.

turned down a lot
Prompt

Here is something I have been thinking about lately: 2023 11 11 11 11 11 11 11 11 11 11

Most negative on

Where the activation went furthest below zero. Blue marks negative activation. Many neurons mean something different in each direction.

  1. #1 peak

    a 1 or 2 doses depends on vaccination history⏎as per Advisory Committee on Immunization Practices annual recommendations on⏎prevention and control of influenza with vaccines.⏎“-” indicates information is not applicable⏎In children, the needle size may

  2. #2 peak

    need more Conjugate Acids and Bases practice, you can also practice Conjugate Acids and Bases practice problems.⏎What professor is this problem relevant for?⏎Based on our data, we think this problem is relevant for Professor Finn

  3. #3 peak

    wide range of data sources, as some sources are complementary and their combined use can result in the production of more detailed and comprehensive statistics.⏎13.15. The integration of different sources and the reuse of surveys for several purposes could also reduce

  4. #4 peak

    is made from sand, soda ash and limestone. Other kinds of glass can contain a variety of other ingredients, which could cause problems in a glass container melting furnace.⏎

Connections

Show on map

Which neurons in earlier layers feed this one, and which later ones it feeds. Amber links excite, blue links suppress; longer bars are stronger.

Fed by earlier layers

  1. L14:238 excites
  2. L2:75 excites
  3. L11:2789 suppresses
  4. L9:25 excites
  5. L1:1 excites
  6. L2:126 suppresses
  7. L6:599 excites
  8. L9:632 excites
  9. L2:1494 excites
  10. L9:159 weakly suppresses
  11. L14:564 weakly excites
  12. L14:116 weakly excites
  13. L1:1029 weakly suppresses
  14. L14:1027 weakly excites
  15. L8:32 weakly excites
  16. L5:158 weakly excites
  17. L15:2852 weakly excites
  18. L11:1579 weakly excites
  19. L2:70 weakly suppresses
  20. L5:3 weakly suppresses
  21. L0:1301 weakly excites
  22. L15:603 weakly suppresses
  23. L0:80 weakly suppresses
  24. L1:1340 weakly excites

Feeds later layers

  1. L20:699 suppresses
  2. L17:50 excites
  3. L17:68 excites
  4. L22:213 suppresses
  5. L18:1075 suppresses
  6. L18:2501 weakly suppresses
  7. L24:531 weakly suppresses
  8. L20:1705 weakly suppresses
  9. L17:644 weakly suppresses
  10. L24:572 weakly suppresses
  11. L21:412 weakly suppresses
  12. L19:2591 weakly suppresses
  13. L18:1179 weakly suppresses
  14. L22:2160 weakly suppresses
  15. L24:20 weakly excites
  16. L25:1529 weakly suppresses
  17. L17:496 weakly excites
  18. L19:2144 weakly suppresses
  19. L18:1185 weakly excites
  20. L17:2155 weakly suppresses
  21. L17:426 weakly excites
  22. L17:336 weakly suppresses
  23. L17:1281 weakly suppresses
  24. L27:2486 weakly suppresses

Activation histogram

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.

1 values between -4.37 and -4.143 values between -3.92 and -3.694 values between -3.69 and -3.466 values between -3.46 and -3.2310 values between -3.23 and -3.0110 values between -3.01 and -2.7819 values between -2.78 and -2.5526 values between -2.55 and -2.3247 values between -2.32 and -2.1058 values between -2.10 and -1.8772 values between -1.87 and -1.64105 values between -1.64 and -1.41174 values between -1.41 and -1.19250 values between -1.19 and -0.96361 values between -0.96 and -0.73561 values between -0.73 and -0.51677 values between -0.51 and -0.28690 values between -0.28 and -0.05356 values between -0.05 and 0.18148 values between 0.18 and 0.4187 values between 0.41 and 0.6349 values between 0.63 and 0.8644 values between 0.86 and 1.0932 values between 1.09 and 1.3227 values between 1.32 and 1.5417 values between 1.54 and 1.7716 values between 1.77 and 2.0011 values between 2.00 and 2.2310 values between 2.23 and 2.4514 values between 2.45 and 2.683 values between 2.68 and 2.914 values between 2.91 and 3.135 values between 3.13 and 3.363 values between 3.36 and 3.595 values between 3.59 and 3.821 values between 3.82 and 4.041 values between 4.27 and 4.501 values between 4.50 and 4.73offnegativepositiveoften

Logit lens

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 layers instead.

Promotes

  1. ني
  2. ’t
  3. mes
  4. ت
  5. :index
  6. DRAW
  7. fluence
  8. 링
  9. enced
  10. SAY

Suppresses

  1. ...";⏎
  2. ;"⏎
  3. ()"⏎
  4. ..."⏎⏎
  5. .";⏎
  6. =";⏎
  7. !";⏎
  8. ."; ⏎
  9. ..."⏎
  10. ".⏎