Neuron Explorer

Qwen3-0.6B

'and' joining paired words

A neuron in layer 20 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 'and' when it links two parallel words or actions, like 'trouble and strife' or 'read and write'. It goes strongly negative on 'but' and on 'and' in a list of steps. Turning it up made the model repeat 'and' many times.

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

    wonder how to subtract mixed numbers? Do you know how to handle unlike denominators or how to regroup or to borrow with mixed numbers? Take a deep breath and let us take it one step at a time. Click and begin.⏎GCF

  2. #2 peak

    are nearby? What type of boundary are the quakes on?⏎Your paper should meet the following requirements:⏎3-4 pages in length (does not include title and reference page)⏎1-2 outside sources⏎Why Work with CheggAnswers⏎

  3. #3 peak

    ⏎Screw Conveyors designs and manufacturer in Coimbatore⏎Screw Conveyors designs and manufacturer in Kerala⏎Screw Conveyors designs and manufacturer in Tamil Nadu⏎Screw Conveyors suppliers in Coimbatore

  4. #4 peak

    here is what I’ve learned along the way:⏎1. Encourage discourse.⏎It may feel a little contrived, but I’ve found that providing discourse stems and sample questions helps students figure out how to leave and respond to comments.⏎2.

  5. #5 peak

    flinching, and insists on pursuing such joy as he can find.⏎Resilience is his middle name.⏎Here circumstances are presented as clouds that help trouble and strife to germinate and proliferate.⏎‘This is the world, and we

  6. #6 peak

    illing’ advisedly because he says it really brought home to him just how fragile the Earth is. “It really brought it home to us that you cannot go and live on Mars, you can’t go and live on Titan, you can’t

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: it so so so so so so so so so so so so so so so so so so so so so so so so so so so so so so so so so so so so so so so so so so so so so so so

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

    instance, if you are thinking of the number 1, it would be "A". Number 2 would be "B". 3 is "C", and so on.⏎- Now, think of a country that starts with the letter you

  2. #2 peak

    ternut Squash can do for you⏎It’s a great hydrator. One serving of butternut squash is roughly 87 percent water, which can help keep you

  3. #3 peak

    and is known only to the Butonian people on the Pulau Makassar. The study used the symbolic interaction

  4. #4 peak

    grams of dietary fiber⏎- 3 grams of sugar⏎- 1.4 grams of protein⏎- 6 milligrams of sodium⏎One serving of butternut squash for vitamins and nutrients is filled with:⏎- Over 10

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. L13:131 strongly excites 'and' joining parallel phrases
  2. L16:376 suppresses 'or' joining alternatives
  3. L4:57 suppresses
  4. L9:92 suppresses "and" or "or" joining two items
  5. L10:74 excites
  6. L18:868 suppresses 'and' joining two nouns in a pair
  7. L15:727 excites
  8. L19:2080 suppresses ampersand joining two items
  9. L1:1250 suppresses
  10. L0:313 excites "and" joining paired nouns
  11. L18:2345 suppresses
  12. L19:547 weakly suppresses
  13. L4:108 weakly excites "or" offering a choice in web text
  14. L9:2010 weakly suppresses
  15. L17:572 weakly excites
  16. L12:51 weakly excites
  17. L4:659 weakly excites
  18. L0:1656 weakly suppresses
  19. L14:154 weakly suppresses
  20. L11:406 weakly suppresses
  21. L18:1107 weakly excites
  22. L5:548 weakly excites
  23. L17:2113 weakly excites
  24. L17:1637 weakly excites

Feeds later layers

  1. L26:347 excites the 're' in self-reliant and similar
  2. L22:632 excites
  3. L25:6 suppresses
  4. L24:205 excites
  5. L21:478 excites
  6. L25:1694 suppresses 'and' joining paired names or fields
  7. L27:568 excites
  8. L23:379 excites
  9. L22:198 suppresses
  10. L27:289 excites
  11. L21:2949 suppresses comma closing 'the more X, the more Y' clauses
  12. L23:2015 suppresses
  13. L26:693 suppresses
  14. L26:65 suppresses
  15. L25:120 suppresses
  16. L24:398 suppresses
  17. L25:1008 suppresses
  18. L21:739 weakly excites 'both X and Y' pairing words
  19. L25:247 weakly suppresses
  20. L21:191 weakly excites
  21. L22:689 weakly suppresses comma closing an intro clause
  22. L26:1232 weakly excites
  23. L26:1007 weakly suppresses
  24. L27:948 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 -41.53 and -40.351 values between -37.99 and -36.811 values between -35.63 and -34.451 values between -33.27 and -32.091 values between -27.37 and -26.191 values between -26.19 and -25.011 values between -25.01 and -23.831 values between -22.65 and -21.474 values between -21.47 and -20.293 values between -20.29 and -19.112 values between -19.11 and -17.933 values between -17.93 and -16.752 values between -16.75 and -15.571 values between -15.57 and -14.395 values between -14.39 and -13.217 values between -13.21 and -12.036 values between -12.03 and -10.854 values between -10.85 and -9.6710 values between -9.67 and -8.4910 values between -8.49 and -7.3113 values between -7.31 and -6.1315 values between -6.13 and -4.9524 values between -4.95 and -3.7722 values between -3.77 and -2.5936 values between -2.59 and -1.4193 values between -1.41 and -0.233455 values between -0.23 and 0.95181 values between 0.95 and 2.133 values between 2.13 and 3.311 values between 4.49 and 5.67offnegativepositiveoften

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. but
  2. 性和
  3. 但
  4. But
  5. but
  6. 但现在
  7. 但他
  8. 但他们
  9. 和支持
  10. 但仍

Suppresses

  1. vice
  2. /or
  3. IGIN
  4. vice
  5. /OR
  6. rog
  7. amp
  8. 蔼
  9. rew
  10. yne