Neuron Explorer

Qwen3-0.6B

commas inside lists of items

A neuron in layer 3 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 fires on a comma in the middle of a list of things, typically just before the last item or "and". Turning it up did not produce any clear theme in the generated text.

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

    . Some of the activities include making a jester hat, learning the chivalric code, making a shield, making a trebuchet, and more.⏎- Knowledge Box Central families: Use this alongside the Middle Ages lap

  2. #2 peak

    fine motor skills to cut out small numbers, think about how to solve six types of math problems, judge the difficulty level and their own capabilities, develop spatial awareness, and learn to apply the correct amount of glue (another fine motor skill, and

  3. #3 peak

    Voyant Tools: Voyant Tools is a free web based suite of tools that allows for you to copy and paste text into a textbox, upload a file, or use an already predetermined corpus and use different tools to visualize your text. The

  4. #4 peak

    Teach them that a good person chooses to think good thoughts, not to say profaned words, do good deeds, go to good places, hang with good people, and find reasons not to fight. Remember your actions speak louder than your words.⏎

  5. #5 peak

    of interest to you.⏎In this guide, we’ll go over all of this in detail. We’ll explore the different types of cobots, their benefits, and how they compare to industrial robots so that you can become better acquainted with this

  6. #6 peak

    Keywords: viral marketing, word-of-mouth marketing, consumption patterns, information traits, credibility.⏎ANTHEUNIS, M.L.,

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: the first person in the room is the first person in the room, the first person in the room, the first person in the room, the first person in the room, the first person in the room, the first person in the room

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

    HGUOT ; reverse order⏎mov char, al⏎L1: mov eax,32 ; creates ASCII code decrement by 32⏎mov char, AL⏎mov al 5-character⏎mov al, TYPE BYTE ; 6

  2. #2 peak

    populations changed over time.⏎Further studies involving more infants could eventually help researchers understand the causes of various intestinal problems that afflict preemies, in particular the sometimes fatal necrotizing enterocolitis, according to researchers at the University of California, Berkeley,

  3. #3 peak

    months in the intensive care unit before they're deemed healthy enough to go home. Between 5 and 10 percent of these preemies develop symptoms of necrotizing enterocolitis (NEC), which requires rounds of antibiotics to halt,

  4. #4 peak

    "⏎- Brix Orolfo, age 12yrs.⏎- © Brix Orolfo⏎Winner

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. L2:66 excites
  2. L2:284 suppresses
  3. L0:2921 suppresses
  4. L2:527 weakly excites
  5. L1:86 weakly excites
  6. L2:36 weakly excites
  7. L0:160 weakly suppresses
  8. L0:2838 weakly excites
  9. L0:1596 weakly excites
  10. L2:2202 weakly excites
  11. L2:2784 weakly suppresses
  12. L0:1756 weakly suppresses
  13. L0:1030 weakly excites
  14. L2:1695 weakly excites
  15. L0:178 weakly suppresses
  16. L2:49 weakly suppresses
  17. L2:1879 weakly suppresses
  18. L1:1029 weakly excites
  19. L0:1119 weakly suppresses
  20. L2:75 weakly suppresses
  21. L1:1975 weakly suppresses
  22. L0:2007 weakly excites 'key' after 'The'
  23. L0:2247 weakly suppresses
  24. L0:1511 weakly suppresses

Feeds later layers

  1. L22:681 excites
  2. L19:2732 weakly suppresses
  3. L8:454 weakly excites dash starting a new list item
  4. L11:2594 weakly suppresses
  5. L9:2666 weakly suppresses
  6. L5:1019 weakly suppresses
  7. L15:2268 weakly suppresses
  8. L4:44 weakly excites
  9. L14:39 weakly suppresses
  10. L20:1752 weakly excites
  11. L8:1460 weakly suppresses
  12. L4:281 weakly suppresses
  13. L17:1287 weakly suppresses
  14. L7:237 weakly excites
  15. L7:2233 weakly excites
  16. L25:524 weakly suppresses
  17. L11:1468 weakly suppresses
  18. L5:2400 weakly suppresses
  19. L14:389 weakly suppresses
  20. L26:106 weakly suppresses
  21. L9:1248 weakly suppresses
  22. L4:96 weakly suppresses
  23. L15:2677 weakly suppresses
  24. L7:1055 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 -1.89 and -1.751 values between -1.61 and -1.484 values between -1.20 and -1.072 values between -1.07 and -0.933 values between -0.93 and -0.802 values between -0.80 and -0.6611 values between -0.66 and -0.5211 values between -0.52 and -0.3978 values between -0.39 and -0.25544 values between -0.25 and -0.111745 values between -0.11 and 0.021187 values between 0.02 and 0.16245 values between 0.16 and 0.2940 values between 0.29 and 0.4312 values between 0.43 and 0.573 values between 0.57 and 0.701 values between 0.70 and 0.844 values between 0.98 and 1.112 values between 1.11 and 1.253 values between 1.38 and 1.522 values between 1.52 and 1.661 values between 1.66 and 1.791 values between 1.79 and 1.931 values between 2.07 and 2.202 values between 2.75 and 2.881 values between 3.16 and 3.291 values between 3.43 and 3.56offnegativepositiveoften

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. engers
  2. 惮
  3. yd
  4. erokee
  5. heel
  6. ilarity
  7. .default
  8. 发展格局
  9. ILES
  10. ulum

Suppresses

  1. .Access
  2. .ACCESS
  3. 手
  4. .access
  5. tain
  6. เหร
  7. 本报
  8. themselves
  9. .setColumns
  10. ammonia