Neuron Explorer

Qwen3-0.6B

'of' after a noun

A neuron in layer 17 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 word 'of' when it follows a noun and is about to name what that noun belongs to, as in 'the size of', 'the length of' or 'the native language of'. When turned up or down, the model drifted into repeated phrases and percentages, which does not clearly match the trigger.

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

    transistor radio, because it’s starting to look a lot like the 1950s.⏎Unfortunately, this won’t be the nostalgic ’50s of big cars and pop music.⏎The 1950s that could be

  2. #2 peak

    consider the following aspects.⏎- Spelling – Convention⏎The spelling of Chinese words using the Latin alphabet (romanisation) seems to be more art than

  3. #3 peak

    the size of a similar-size hoofed mammal and eyes only a third as large.⏎The

  4. #4 peak

    technical documentation. However, it is often not the native language of the readers (or even of the authors) of such documentation. Many readers have

  5. #5 peak

    monopoly on the newly discovered region. He never waiver in his belief that he had found an alternate route to Asia. Columbus had found a paradise just not the one of his hopes and aspirations. This paradise was right before his eyes, yet he lost

  6. #6 peak

    of the most technically demanding ever written for the piano, he is known more for his nuance, his expressive depth and his ability to conjure up the melody of the human voice from the instrument’s keys. For this sublime blend of skill and

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: what is what, which is itself, to what it do. In this, I am both of what I am and in what I am. In what I am, are there be be be be be be be be be be be be

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

    reports that traffic engineers could use to adjust traffic lights or improve roads.⏎Bill Gates and Paul Allen were high school students at Lakeside School in Seattle. The Lakeside Programmers Group got free

  2. #2 peak

    From Tree Roots⏎- Locate sewer lines – You can call your local public works office or the national 811 “Call Before You Dig” number to locate underground utilities on your property. Before

  3. #3 peak

    perfect companion for study and revision for nursing and healthcare students from the publishers of the market-leading at a Glance series. This easy-to-read, accessible guide brings together all the key principles of caring for patients with medical and

  4. #4 peak

    and offers meditative benefits.”⏎This demand fits with the overall interest he’s seeing among clients for less intensive fitness routines that are accessible and mesh with an on-the-go lifestyle.⏎Why gardening is exercise⏎Gardening engages all the major muscle groups

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. L15:888 strongly excites
  2. L16:1594 strongly excites
  3. L16:652 strongly suppresses
  4. L14:935 excites
  5. L9:2661 excites
  6. L11:2859 suppresses
  7. L10:321 suppresses
  8. L10:365 suppresses
  9. L5:493 excites
  10. L15:1888 excites
  11. L10:1644 excites
  12. L10:556 suppresses
  13. L9:227 excites
  14. L15:2253 excites
  15. L4:392 suppresses
  16. L1:740 suppresses
  17. L9:662 suppresses
  18. L9:74 suppresses
  19. L10:120 suppresses
  20. L16:513 suppresses
  21. L2:1439 excites
  22. L10:1289 excites
  23. L10:1278 excites
  24. L10:1377 suppresses

Feeds later layers

  1. L18:1782 strongly suppresses
  2. L19:1023 suppresses
  3. L20:438 suppresses
  4. L19:644 suppresses
  5. L19:1180 excites
  6. L19:1539 excites
  7. L18:1041 excites
  8. L21:713 excites
  9. L20:596 suppresses
  10. L21:614 suppresses
  11. L25:284 excites
  12. L20:1883 suppresses 'the' inside a mid-sentence phrase
  13. L23:79 excites
  14. L24:153 excites
  15. L25:544 excites
  16. L18:41 excites
  17. L20:219 suppresses "the" after time or place prepositions
  18. L18:376 excites
  19. L19:2115 excites
  20. L19:2815 suppresses
  21. L21:675 excites
  22. L20:47 suppresses
  23. L20:738 excites
  24. L20:809 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.

2 values between -9.37 and -8.991 values between -8.60 and -8.221 values between -7.08 and -6.701 values between -6.70 and -6.321 values between -6.32 and -5.932 values between -5.93 and -5.555 values between -5.55 and -5.176 values between -5.17 and -4.799 values between -4.79 and -4.4116 values between -4.41 and -4.0321 values between -4.03 and -3.6529 values between -3.65 and -3.2639 values between -3.26 and -2.8858 values between -2.88 and -2.5082 values between -2.50 and -2.12123 values between -2.12 and -1.74186 values between -1.74 and -1.36296 values between -1.36 and -0.98372 values between -0.98 and -0.59580 values between -0.59 and -0.211109 values between -0.21 and 0.17467 values between 0.17 and 0.55210 values between 0.55 and 0.93106 values between 0.93 and 1.3168 values between 1.31 and 1.7042 values between 1.70 and 2.0816 values between 2.08 and 2.4615 values between 2.46 and 2.8415 values between 2.84 and 3.2212 values between 3.22 and 3.609 values between 3.60 and 3.983 values between 3.98 and 4.371 values between 4.37 and 4.754 values between 4.75 and 5.131 values between 5.51 and 5.89offnegativepositiveoften

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. cles
  2. clus
  3. hung
  4. {text
  5. C
  6. Pascal
  7. ikers
  8. 的任务
  9. the
  10. 的看法

Suppresses

  1. 外围
  2. propio
  3. ligt
  4. เสม
  5. educação
  6. liest
  7. ดาร
  8. 绪
  9. -largest
  10. (primary