Neuron Explorer

Qwen3-0.6B

line breaks inside wrapped text

A neuron in layer 15 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 mostly fires on line breaks or the word piece right before one, especially inside poems and wrapped paragraphs. It seems to sense where a line of text is about to end. It tends to go negative right before a list or a colon. Turning it up gave off-topic rambling.

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

  1. #1 peak

    who already have high-incomes; almostnone of it goes to the poorest Americans who need the most helpbuilding the financial assets that can lead to long-term economicopportunities and security.<|endoftext|>Modern methods of time management incorporate various ways of recording time

  2. #2 peak

    -eastern Namibia and South Africa, specifically in the Kruger⏎National Park, KwaZulu-Natal and the Eastern Cape. It prefers woodland dominated⏎by a broad-leaved trees such as bushwillows (Combretum),

  3. #3 peak

    , at sacred-texts.com⏎LONG is the mountain pheasant's tail⏎That curves down in its flight;⏎But longer still, it seems to me,⏎Left in my lonely plight,⏎Is this unending night.⏎The writer was a

  4. #4 peak

    I could not take,⏎Although so well inclined.⏎The rebels came with savage yells,⏎A Yankee they had found,⏎With curses bid me “face about,”⏎And march away to town.⏎They marched me down with many more,⏎Through mud

  5. #5 peak

    deficiency, its clinical manifestations and⏎the possible repercussions for Australia’s refugee population.⏎Full blood count and blood film, iron studies and haemoglobinopathy studies are⏎routinely performed for newly arrived refugees in Australia. At the Migrant Health

  6. #6 peak

    this spring morning, and missed the dawn, From everywhere I heard the cry of birds. That night the sound of wind and rain had come, Who knows how many petals then had fallen?<|endoftext|>At EdPlace we are passionate about

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: I have a but I have I have a This is the same as...? Wait, that's the same... Wait, I'm getting Wait, I'm Wait, I'm I'm Wait Wait

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

    time if the child does not get any help. Some signs a child is having difficulties can be:⏎Behavior problems are usually easy to see. They can include ongoing:⏎- aggression or bullying⏎- refusal to do what they are asked

  2. #2 peak

    -blame/failure/low self-esteem Sadness/depression: loss . In Building Self-Esteem in Children Who Are Adopted or Fostered, Dr. Sue offers simple and practical advice to those supporting children aged 7+ to help

  3. #3 peak

    People can be exposed to RF radiation from both natural and man-made sources. Natural sources include:⏎Man-made RF radiation is used for many different things, such as:⏎Most people are exposed to much lower levels of man-made RF radiation every day due

  4. #4 peak

    Biff should be making good money and blames his son’s failure on his laziness. But it is Willy who has se... ... middle of paper ... ...d a the country”

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. L10:230 suppresses
  2. L9:234 suppresses line break inside hard-wrapped text
  3. L10:2315 excites line break inside wrapped text
  4. L7:239 excites
  5. L10:494 suppresses
  6. L7:133 suppresses line break inside a sentence
  7. L7:1240 suppresses
  8. L13:2091 suppresses
  9. L7:2925 suppresses
  10. L13:309 excites
  11. L9:1238 excites
  12. L10:258 excites
  13. L12:601 weakly excites question mark or line break after an FAQ heading
  14. L12:2216 weakly excites
  15. L9:511 weakly suppresses
  16. L8:2432 weakly suppresses line breaks in lists and wrapped text
  17. L3:79 weakly excites
  18. L8:380 weakly suppresses
  19. L12:1012 weakly excites
  20. L14:2751 weakly suppresses
  21. L9:246 weakly suppresses
  22. L9:161 weakly excites
  23. L9:909 weakly suppresses
  24. L2:140 weakly excites "one" after an ordinal or superlative

Feeds later layers

  1. L17:1242 excites
  2. L16:442 excites
  3. L17:554 excites
  4. L16:822 excites
  5. L16:317 suppresses
  6. L20:1467 weakly suppresses
  7. L17:916 weakly excites
  8. L19:435 weakly suppresses
  9. L24:1197 weakly excites
  10. L25:239 weakly suppresses
  11. L24:406 weakly suppresses
  12. L17:1612 weakly excites
  13. L23:1763 weakly suppresses
  14. L17:1564 weakly excites
  15. L16:2249 weakly suppresses
  16. L17:2056 weakly suppresses
  17. L19:323 weakly suppresses
  18. L19:970 weakly excites
  19. L20:1571 weakly excites
  20. L22:667 weakly suppresses
  21. L17:1342 weakly suppresses
  22. L24:1619 weakly suppresses
  23. L26:144 weakly excites
  24. L17:2875 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 -2.60 and -2.361 values between -2.36 and -2.135 values between -2.13 and -1.892 values between -1.89 and -1.664 values between -1.66 and -1.421 values between -1.42 and -1.197 values between -1.19 and -0.9511 values between -0.95 and -0.7222 values between -0.72 and -0.48224 values between -0.48 and -0.251802 values between -0.25 and -0.011664 values between -0.01 and 0.22151 values between 0.22 and 0.467 values between 0.46 and 0.702 values between 0.70 and 0.931 values between 1.87 and 2.111 values between 2.58 and 2.811 values between 3.05 and 3.281 values between 6.57 and 6.81offnegativepositiveoften

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. .gif
  3. acionales
  4. etration
  5. /releases
  6. igans
  7. �
  8. /loader
  9. .ogg
  10. .bmp

Suppresses

  1. auto
  2. matic
  3. WEST
  4. aut
  5. Auto
  6. InitialState
  7. AUTO
  8. 出来
  9. 提供优质
  10. �