Neuron Explorer

Qwen3-0.6B

line break at the end of a line

A neuron in layer 24 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 responds to the point where a line of text ends and a new line begins. When it is turned up, the model starts writing nothing but blank line breaks, so it is tied to newlines.

Written by Claude from this neuron's activations and steering outputs. Confidence: high. 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

    , delivering a smooth user experience. With the ever-growing demand for innovative apps, proficient coding skills have become invaluable for developers looking to make a mark in the industry.⏎The Challenges of Coding for Apps⏎Coding for apps presents its fair share of challenges

  2. #2 peak

    screen time ) 😌⏎- Go to bed and wake up at the same time each day. Our physical

  3. #3 peak

    Symmetric encryption AES/block ciphers Block modes (ECB, CTR, OFB) Asymmetric encryption RSA in Java Comparison of algorithms Key sizes Hash functions⏎Secure hash functions in Java⏎As you probably know if you've worked on certain

  4. #4 peak

    package can be installed like so:⏎cabal install bool-extras⏎If you don’t have cabal-install, you can manually download and install bool

  5. #5 peak

    ⏎- Configure Spark to modify the Spark properties, environmental variables, or logging properties⏎- Monitor Spark using the web UIs, metrics, and external instrumentation⏎-

  6. #6 peak

    ⏎- Subject: Drama - World⏎- Type: Essay⏎- Paper: #87743533⏎Excerpt from Essay :⏎Perceptions of Che Guevera⏎PERCEPTION

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:fullyfullyfullyfullyfullyfullyfullyfullyfullyfullyfullyfullyfullyfullyfullyfullyfullyfullyfullyfullyfullyfullyfullyfully

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

    The investigation was carried out using AWS 5.22–95 filler wire of size 1.2 mm diameter and the base metal used was IS:2062 structural steel plate of 20 mm thickness. An

  2. #2 peak

    done.⏎Answer: True. There are hundreds of thousands of dust mites in our homes, from

  3. #3 peak

    70%.⏎Answer: B. Humidity levels should be kept below 60 percent, ideally between

  4. #4 peak

    bad mood.⏎- He’s very cross today, so don’t bother him.⏎synonyms: cranky, disagreeable, grouchy⏎Word Builder: cross +⏎crosswalk: a part of the road that is marked off for people to cross

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. L21:173 strongly excites
  2. L15:123 suppresses
  3. L18:48 suppresses
  4. L20:8 suppresses
  5. L14:51 excites colon that introduces a statement
  6. L17:38 suppresses
  7. L19:187 suppresses
  8. L13:10 excites
  9. L9:7 excites
  10. L10:84 suppresses
  11. L17:829 suppresses
  12. L15:44 suppresses
  13. L16:170 suppresses
  14. L20:863 excites
  15. L18:5 suppresses
  16. L12:5 excites
  17. L10:33 suppresses
  18. L14:56 suppresses document ends and list line breaks
  19. L22:262 suppresses comma inside big numbers
  20. L17:62 suppresses
  21. L15:81 suppresses line breaks inside running text
  22. L23:35 excites
  23. L14:348 suppresses
  24. L1:1 excites

Feeds later layers

  1. L27:487 strongly excites
  2. L26:42 strongly suppresses
  3. L26:241 strongly excites
  4. L25:40 strongly suppresses
  5. L25:91 strongly excites
  6. L27:1801 strongly excites
  7. L25:248 strongly suppresses
  8. L26:230 excites
  9. L25:297 excites
  10. L25:533 excites first letter m or oct fragments
  11. L27:143 suppresses
  12. L27:598 suppresses
  13. L26:753 suppresses
  14. L27:1496 excites
  15. L26:1018 excites
  16. L26:156 weakly suppresses
  17. L25:144 weakly excites
  18. L27:116 weakly suppresses
  19. L25:71 weakly suppresses
  20. L26:167 weakly excites
  21. L26:685 weakly suppresses
  22. L27:200 weakly suppresses
  23. L27:1687 weakly excites
  24. L27:433 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 -16.39 and -15.611 values between -15.61 and -14.822 values between -13.25 and -12.471 values between -10.11 and -9.331 values between -9.33 and -8.547 values between -8.54 and -7.761 values between -7.76 and -6.977 values between -6.97 and -6.196 values between -6.19 and -5.408 values between -5.40 and -4.627 values between -4.62 and -3.8312 values between -3.83 and -3.0522 values between -3.05 and -2.2634 values between -2.26 and -1.4887 values between -1.48 and -0.692480 values between -0.69 and 0.091041 values between 0.09 and 0.8867 values between 0.88 and 1.6631 values between 1.66 and 2.4528 values between 2.45 and 3.2315 values between 3.23 and 4.029 values between 4.02 and 4.809 values between 4.80 and 5.593 values between 5.59 and 6.378 values between 6.37 and 7.162 values between 7.16 and 7.945 values between 7.94 and 8.732 values between 8.73 and 9.511 values between 9.51 and 10.305 values between 10.30 and 11.081 values between 11.08 and 11.872 values between 11.87 and 12.651 values between 13.44 and 14.221 values between 14.22 and 15.01offnegativepositiveoften

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. ⏎ ⏎
  5. ⏎ ⏎
  6. ⏎⏎⏎
  7. ⏎ ⏎
  8. ⏎ ⏎
  9. ⏎⏎⏎⏎⏎⏎

Suppresses

  1. ly
  2. ful
  3. LY
  4. ся
  5. gether
  6. shire
  7. theless
  8. alyze
  9. hip
  10. ">'+⏎