Neuron Explorer

Qwen3-0.6B

comma after an opening clause

A neuron in layer 18 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 the comma that ends an introductory phrase such as 'When he arrived,' or 'If I ask for something,' just before the main sentence starts. Turned up, the model fell into repeated phrases like 'I feel, I feel', which does not clearly relate to this trigger.

Written by Claude from this neuron's activations and steering outputs. Confidence: medium. Think it is wrong? Flag it (opens an email to the maker with the neuron already filled in).

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

    found on beaches worldwide. When one thinks of the word recycling, it is newspapers, cardboard, glass, steel, aluminum foil, aluminum cans,

  2. #2 peak

    stick to it.⏎4. Start thinking more positively in order to build self-confidence.⏎How to identify and challenge negative thoughts:⏎When we’re faced with negative thoughts, it can be hard to know where to start. We may find it difficult to

  3. #3 peak

    genuine a blessing that phrase is. Bed bugs can make your life miserable and can be very difficult to get rid of.⏎When most of us think of bed bugs, we think of one species, but did you know there are more than ninety?

  4. #4 peak

    years were over he had said all that he had to say, for his conversation was limited, and he determined to return to his own castle. When he arrived, he saw the children playing in the

  5. #5 peak

    Jesus and his disciples on many occasions. Surely, Jesus would fulfill their request.⏎But that isn’t what happened.⏎When Jesus received word about Lazarus’ condition, he stayed where he was for the next two days. It wasn’t until the

  6. #6 peak

    code "reads" easily, and yet it has the flexibility to contain all the different things I need to return. If I ask for something that isn't there, I get an exception, which is life as usual for a Python programmer.⏎if

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: and then, and then, and then, and then, and then, and then, and then, and then, and then, and then, and then, and then, and then, and then, and then, and then,

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

    any reconfiguration event, there is no need to run this command interactively.⏎For more information, see the following references:<|endoftext|>The United Kingdom, with its constituent parts, including

  2. #2 peak

    up on the Web—until I saw one geek-speak phrase on the home page, so small I almost missed it: API Documentation. To understand its significance, let me tell you how I got subway schedules on my iPhone.⏎Just a few

  3. #3 peak

    link a person’s thoughts, conceptualizations, and memories with a computer. An Imagination Interface, perhaps. By any name, it doesn’t take much time to realize that you have always wanted one

  4. #4 peak

    one time link to download ebooks to your computer. Orders paid by PayPal require up to 8 business hours to verify payment and release electronic media. For immediate downloads, payment with credit card is required.⏎This math unit presents four activity pages in which

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. L13:1232 excites
  2. L17:1426 excites
  3. L16:1886 excites
  4. L4:2769 suppresses
  5. L17:1845 excites
  6. L11:1084 excites
  7. L8:410 suppresses
  8. L9:1579 suppresses
  9. L13:530 suppresses commas after a 'since' clause
  10. L14:1209 suppresses
  11. L15:2295 suppresses
  12. L13:706 excites
  13. L17:3021 excites
  14. L6:1485 excites
  15. L15:1830 suppresses
  16. L6:2233 excites
  17. L17:2397 excites
  18. L13:1273 suppresses
  19. L16:2210 suppresses
  20. L13:921 weakly excites
  21. L6:2843 weakly suppresses
  22. L14:2927 weakly suppresses comma after a conditional lead-in clause
  23. L17:2960 weakly suppresses
  24. L16:968 weakly excites

Feeds later layers

  1. L22:1551 weakly excites
  2. L20:2891 weakly excites
  3. L21:1691 weakly excites "at" in "at first" and "initial" words
  4. L27:910 weakly excites
  5. L20:2484 weakly suppresses
  6. L22:689 weakly excites comma closing an intro clause
  7. L24:33 weakly suppresses
  8. L19:2726 weakly suppresses
  9. L20:501 weakly suppresses
  10. L24:13 weakly suppresses
  11. L19:2508 weakly excites
  12. L21:2941 weakly suppresses
  13. L22:2314 weakly excites
  14. L19:2440 weakly suppresses
  15. L20:1281 weakly suppresses
  16. L21:1044 weakly excites
  17. L23:1413 weakly excites
  18. L20:610 weakly suppresses
  19. L25:299 weakly suppresses
  20. L19:2023 weakly suppresses
  21. L21:1612 weakly excites
  22. L21:2901 weakly suppresses
  23. L22:1391 weakly excites
  24. L24:945 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.

2 values between -4.01 and -3.622 values between -2.83 and -2.443 values between -2.44 and -2.053 values between -2.05 and -1.653 values between -1.65 and -1.2610 values between -1.26 and -0.87105 values between -0.87 and -0.471529 values between -0.47 and -0.082023 values between -0.08 and 0.31175 values between 0.31 and 0.7119 values between 0.71 and 1.106 values between 1.10 and 1.499 values between 1.49 and 1.886 values between 1.88 and 2.282 values between 2.28 and 2.672 values between 3.06 and 3.461 values between 3.46 and 3.851 values between 3.85 and 4.241 values between 4.64 and 5.032 values between 5.03 and 5.421 values between 6.99 and 7.391 values between 7.39 and 7.781 values between 9.35 and 9.751 values between 11.32 and 11.71offnegativepositiveoften

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. 首先
  2. 首先要
  3. INGS
  4. 最先
  5. zunächst
  6. 务必
  7. liğinde
  8. usting
  9. aturas
  10. 必定

Suppresses

  1. その後
  2. [str
  3. _refer
  4. ضاف
  5. InSection
  6. そして
  7. onnen
  8. ��
  9. luyện
  10. roduce