Neuron Explorer

Qwen3-0.6B

semicolons and list-like citations

A neuron in layer 23 of Qwen3-0.6B. Its strongest snippets are mixed: they peak on different tokens. It is featured, so it has steering outputs.

What this neuron does

This neuron responds to places where a semicolon or a list of items is likely to come next, and to some bits of catalogue or citation text. Turning it up made the model output a long string of semicolons.

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

    Show all Grade Level⏎Resource Type: Activities⏎Results 1 - 7 of 7 matches⏎From Grid to Home⏎Lisa Gardiner, National Center for Atmospheric Research; Marie Johnson, United States Military Academy; Jonathan Hoffman, University of

  2. #2 peak

    you fold⏎This problem focuses on Dienes' Logiblocs. What is the same and⏎what is different about these pairs of shapes? Can you describe the⏎shapes in the picture?⏎10 space travellers are waiting to board their spaces

  3. #3 peak

    ology and healthcare accessibility. Formulation of regional guidelines may help reduce AOM burden.⏎- acute otitis media⏎Statistics from Altmetric.com⏎If you wish to reuse any or all of this article please use the link below which will take you

  4. #4 peak

    God and is profitable for doctrine, for reproof, for correction, for instruction in righteousness: That the man of God may be perfect, thoroughly

  5. #5 peak

    usev crater<|endoftext|>It is really amazing how Titanic still captures our imagination 100 years after she sunk to the bottom of the Atlantic in 1912.⏎I put this down to the fact that Titanic is firstly an amazing tale

  6. #6 peak

    James McIntyre⏎Read this poem in other languages⏎This poem has not been translated into any other language yet.⏎Still I Rise⏎The Road Not Taken⏎If You Forget Me⏎Edgar Allan Poe⏎Stopping By Woods On A Snowy

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: there is a type of research that research at the research and research research research research research research research research research research research research research research research research research research research research research research research research research research research research research research research research research research research research

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

    Vespidae) are widely distributed around most of the globe and are diverse and common in many landscapes (Beggs et al., 2011). Paper wasps are likely to influence many other species in terrestrial ecosystems because they

  2. #2 peak

    .⏎from the GNU version of the Collaborative International Dictionary of English⏎- n. Use; practice; exercise.⏎- transitive v. To use; to exercise; to inure; to accustom by practice.⏎from The Century Dictionary and

  3. #3 peak

    , 286,028; Lakewood, 143,754; Fort Collins, 124,665; Westminster, 103,599; and Pueblo,

  4. #4 peak

    comment.<|endoftext|>100 Chemistry Questions and Answers 8211; Basic Chemistrynbsp;nbsp;Chemistry Questions 8211; Learn Basic Chemistry Part 1 (1-25)nbsp;1

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. L20:373 strongly excites
  2. L21:351 suppresses
  3. L19:1763 excites
  4. L20:206 suppresses
  5. L16:1598 excites
  6. L17:2027 excites
  7. L18:1465 excites semicolon joining clauses
  8. L19:2352 excites
  9. L22:381 weakly suppresses
  10. L4:2749 weakly excites
  11. L20:2986 weakly excites
  12. L17:1134 weakly excites
  13. L19:1020 weakly excites
  14. L10:1982 weakly suppresses
  15. L21:881 weakly suppresses comma after 'if' clause in prose, braces in code
  16. L2:1164 weakly excites
  17. L22:2647 weakly excites
  18. L21:1222 weakly suppresses
  19. L14:2182 weakly excites
  20. L21:2735 weakly suppresses "it" after whether or when
  21. L14:1790 weakly excites
  22. L13:2085 weakly excites
  23. L22:2098 weakly suppresses
  24. L5:2200 weakly suppresses

Feeds later layers

  1. L24:378 strongly excites
  2. L27:347 excites
  3. L26:421 excites
  4. L25:2637 excites
  5. L24:264 excites
  6. L25:318 excites
  7. L25:832 excites zero before a decimal point
  8. L25:296 weakly excites
  9. L26:603 weakly excites
  10. L26:1329 weakly excites
  11. L24:2812 weakly suppresses
  12. L24:2381 weakly excites
  13. L25:238 weakly excites
  14. L24:3025 weakly suppresses
  15. L25:511 weakly suppresses
  16. L25:461 weakly suppresses
  17. L24:2194 weakly excites
  18. L24:531 weakly excites
  19. L27:1985 weakly suppresses
  20. L27:1408 weakly suppresses
  21. L25:2289 weakly suppresses
  22. L27:634 weakly excites
  23. L24:962 weakly suppresses
  24. L25:2743 weakly excites

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 -10.50 and -10.111 values between -6.96 and -6.561 values between -4.59 and -4.201 values between -3.81 and -3.411 values between -3.41 and -3.021 values between -3.02 and -2.631 values between -2.63 and -2.231 values between -2.23 and -1.842 values between -1.84 and -1.447 values between -1.44 and -1.0530 values between -1.05 and -0.66259 values between -0.66 and -0.262308 values between -0.26 and 0.131129 values between 0.13 and 0.53141 values between 0.53 and 0.9216 values between 0.92 and 1.312 values between 1.31 and 1.714 values between 1.71 and 2.101 values between 2.49 and 2.891 values between 4.86 and 5.25offnegativepositiveoften

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. ;
  2. ;
  3. ;
  4. ;b
  5. .;
  6. ;⏎
  7. ;c
  8. ;",
  9. ;"
  10. ;a

Suppresses

  1. ughter
  2. Matcher
  3. عد
  4. matches
  5. 才发现
  6. ####⏎
  7. .matches
  8. matches
  9. 等待
  10. Matching