Neuron Explorer

Qwen3-0.6B

'of' inside title-like noun phrases

A neuron in layer 8 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

The neuron fires on the word 'of' when it links two nouns in a set phrase or title, such as 'Sunday of Sundays', 'Harvest of Memories', or 'Copyright of the Federal'. The steering samples do not show a clear effect, so the pattern is known mainly from the snippets.

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

    on Sunday, the Lord’s Day), but also in a special festival of the resurrection. This festival we call Easter Day, and it is celebrated as the Sunday of Sundays!⏎Lent, as a season of preparation, is traditionally focused on repent

  2. #2 peak

    are twice as likely as whites to have had 10 or more spells of unemployment over their prime working years.⏎-<|endoftext|>In an ideal world, a material that is inherently resistant to its service environment complies with the mechanical, formability and

  3. #3 peak

    4th of July should be a time when we come together to honor our country by celebrating our

  4. #4 peak

    Ewan Boyce and later the property of her husband Captain William Boyce<|endoftext|>A Point of View: Making friends the shared currency way⏎- 2 March 2012⏎- From the section Magazine⏎Greece is falling out

  5. #5 peak

    lected After Runoff Victory|⏎Federal Elections (94-2004): Election Results for the US Senate⏎and the US House of Representatives. Copyright of the Federal⏎Elections

  6. #6 peak

    - Get a towel (10 seconds)⏎- Answer the phone (2 minutes)⏎- Answer the front door (4-6 minutes)⏎ABC & Ds of Drowning Prevention⏎Read the following ABC and Ds of drowning prevention to learn

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: how do you know what you think? And what do you think about? What do you think about? What do you think about? What do you think about? What do you think about? What do you think about? What do you

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

    lectronics has not been given much consideration previously, though much effort has been devoted to the development of general-purpose nanoelectronic systems, i.e

  2. #2 peak

    ptic. They are used traditionally to treat headaches[⏎If you have any useful information about this plant, please leave a comment. Comments have to be approved before they are shown here.<|endoftext|>Save water with wind power⏎21 March 20

  3. #3 peak

    as the behaviour of those in our local community.⏎Cultural – Listening to and incorporating elements of<|endoftext|>- A book listing the names, addresses, and telephone numbers of the people in a particular area⏎- means a Telephone Directory published by BT or

  4. #4 peak

    journalistic institution, Adafruit strives to be a fair, informative, and positive voice within the community – check it out here: adafruit.com/editor

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. L5:5 suppresses digit after a label like paper or type
  2. L4:290 excites
  3. L4:4 suppresses
  4. L6:58 excites
  5. L7:195 suppresses 'of' in set noun phrases
  6. L6:43 excites
  7. L6:136 suppresses
  8. L4:250 excites
  9. L4:1300 weakly suppresses
  10. L1:978 weakly excites
  11. L3:941 weakly suppresses
  12. L4:315 weakly suppresses
  13. L6:10 weakly excites
  14. L7:1759 weakly excites
  15. L3:315 weakly excites 'of' after a noun phrase
  16. L3:786 weakly excites
  17. L3:72 weakly excites
  18. L6:60 weakly suppresses
  19. L7:56 weakly excites
  20. L7:1760 weakly suppresses
  21. L7:386 weakly suppresses
  22. L4:493 weakly excites
  23. L5:509 weakly excites the word 'even' after a clause
  24. L0:2535 weakly suppresses

Feeds later layers

  1. L10:959 excites
  2. L10:319 suppresses
  3. L10:121 suppresses
  4. L10:502 excites
  5. L9:976 suppresses
  6. L10:152 suppresses
  7. L10:67 suppresses
  8. L18:67 weakly excites unclear: mixed mid-sentence words
  9. L9:290 weakly suppresses
  10. L9:708 weakly excites
  11. L15:1101 weakly excites
  12. L9:61 weakly suppresses
  13. L14:251 weakly suppresses
  14. L12:99 weakly excites
  15. L9:742 weakly suppresses "of" after an agency or department name
  16. L15:795 weakly excites
  17. L15:638 weakly suppresses "the" inside institution names
  18. L11:453 weakly suppresses
  19. L16:689 weakly suppresses 'the' in fixed phrases like 'ahead of the curve'
  20. L9:172 weakly excites
  21. L20:146 weakly excites
  22. L10:29 weakly suppresses
  23. L15:1580 weakly suppresses
  24. L14:2349 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 -1.42 and -1.283 values between -1.28 and -1.144 values between -1.14 and -1.0010 values between -1.00 and -0.8611 values between -0.86 and -0.7219 values between -0.72 and -0.5844 values between -0.58 and -0.44108 values between -0.44 and -0.30384 values between -0.30 and -0.161196 values between -0.16 and -0.021466 values between -0.02 and 0.12324 values between 0.12 and 0.26118 values between 0.26 and 0.4060 values between 0.40 and 0.5436 values between 0.54 and 0.6825 values between 0.68 and 0.8221 values between 0.82 and 0.9615 values between 0.96 and 1.1018 values between 1.10 and 1.248 values between 1.24 and 1.386 values between 1.38 and 1.525 values between 1.52 and 1.665 values between 1.66 and 1.805 values between 1.80 and 1.955 values between 1.95 and 2.094 values between 2.09 and 2.232 values between 2.23 and 2.372 values between 2.37 and 2.511 values between 2.79 and 2.932 values between 4.05 and 4.19offnegativepositiveoften

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. ately
  2. adamente
  3. weise
  4. /span
  5. eward
  6. /style
  7. atement
  8. emachine
  9. 'em
  10. �回

Suppresses

  1. t
  2. d
  3. фор
  4. 领域
  5. ,“
  6. ĩnh
  7. pé
  8. ích
  9. powerful
  10. 的能力