Neuron Explorer

Qwen3-0.6B

'of' after a page, article or section number

A neuron in layer 5 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 comes right after a number that names a page, article, section or chapter, such as 'page 17 of the textbook' or 'Article 123 of the Constitution'. Turning it up didn't make the text about any particular subject.

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

    carriage return (no need to deal with errors, etc.).⏎Reading a character from the keyboard is done via the read_char macro (see page 17 of the textbook for a description). So you'll have to call this macro 5

  2. #2 peak

    years ago— which the new figures now make out-of-date. (The state’s estimates of sea-level rise and subsidence are listed on page 83 of the Master Plan.)⏎The state plan, while “valuable and thoughtful,” has

  3. #3 peak

    These holidays are defined in article 18 of the Afghan constitution from 2004. The holidays occur on a fixed

  4. #4 peak

    ? [Hint: remove any blanks from each column, remembering to sort the columns together.] Use an F test (on pages A-2 and A-3 of your Class Notes book). (You are advised NOT to use the Two-sample F

  5. #5 peak

    .⏎We need to grumble and complain about drivers who fail to use good winter tires and who overdrive the conditions.<|endoftext|>Viewing page 28 of 101⏎It looks like you're using a mobile device. We

  6. #6 peak

    country and for restoring Mexico to a Republic.⏎May 1st: Labor Day, commemorates the union movement of the Mexican worker. Article 123 of the Constitution of 1917 gave workers the right to organize labor unions

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: I need to get this problem to be. So,, I need to get this problem to be,, I need to get,,, and the, and, and others, and the, and, and others, and the,

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

    played in time (that is what rhythm is!!).⏎Many people count sixteenth notes this way: ‘one ee and a, two ee and a, three ee

  2. #2 peak

    Start with the sentence ‘Like the Corelli, the harmony in Haydn’s Joke quartet is mainly diatonic and functional but with some occasional dissonance’.⏎No homework this week⏎For Friday 12th – Finish

  3. #3 peak

    last bar, each note is played on each beat.⏎There are a few different ways of counting eighth and sixteenth notes.⏎For eighths notes you can count ‘1 and 2 and

  4. #4 peak

    hour.⏎Pythagoras' concepts included the mathematical order of the cosmos, and he may have been led to this assessment from the mathematical order of music (consonants of octaves, fifths, and fourths being produced by simple ratios

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. L3:2441 excites
  2. L4:2255 weakly suppresses
  3. L0:1450 weakly excites
  4. L3:2928 weakly excites
  5. L4:1140 weakly excites
  6. L2:521 weakly suppresses
  7. L4:906 weakly excites
  8. L3:797 weakly excites
  9. L1:2061 weakly suppresses
  10. L3:2655 weakly excites
  11. L1:1226 weakly excites
  12. L0:1545 weakly excites
  13. L3:616 weakly suppresses
  14. L1:2084 weakly suppresses
  15. L3:856 weakly suppresses
  16. L2:112 weakly excites
  17. L3:286 weakly excites
  18. L4:2697 weakly suppresses
  19. L0:392 weakly suppresses
  20. L1:3071 weakly suppresses
  21. L3:72 weakly suppresses
  22. L2:59 weakly excites
  23. L3:1135 weakly excites
  24. L2:996 weakly excites

Feeds later layers

  1. L9:602 excites
  2. L12:752 excites
  3. L16:3051 weakly suppresses
  4. L9:673 weakly excites
  5. L7:1941 weakly excites
  6. L20:2672 weakly suppresses
  7. L9:942 weakly excites
  8. L6:336 weakly excites
  9. L6:998 weakly suppresses
  10. L6:345 weakly excites
  11. L10:1626 weakly excites
  12. L12:203 weakly excites
  13. L14:629 weakly suppresses
  14. L21:893 weakly excites
  15. L21:1702 weakly suppresses
  16. L20:1236 weakly excites
  17. L13:1925 weakly suppresses
  18. L18:925 weakly suppresses
  19. L17:2109 weakly suppresses
  20. L12:1594 weakly suppresses
  21. L7:2784 weakly suppresses unclear: 'Match' in exercise instructions
  22. L6:2217 weakly excites
  23. L11:1147 weakly excites
  24. L9:926 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 -0.63 and -0.581 values between -0.58 and -0.533 values between -0.53 and -0.485 values between -0.48 and -0.4412 values between -0.44 and -0.3930 values between -0.39 and -0.3448 values between -0.34 and -0.29100 values between -0.29 and -0.24159 values between -0.24 and -0.20265 values between -0.20 and -0.15360 values between -0.15 and -0.10529 values between -0.10 and -0.05608 values between -0.05 and -0.01663 values between -0.01 and 0.04455 values between 0.04 and 0.09294 values between 0.09 and 0.14191 values between 0.14 and 0.1990 values between 0.19 and 0.2346 values between 0.23 and 0.2818 values between 0.28 and 0.3312 values between 0.33 and 0.387 values between 0.38 and 0.432 values between 0.43 and 0.473 values between 0.47 and 0.521 values between 0.52 and 0.572 values between 0.57 and 0.621 values between 0.86 and 0.901 values between 1.19 and 1.241 values between 1.24 and 1.29offnegativepositiveoften

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. kip
  2. kop
  3. ultipartFile
  4. overall
  5. 的整体
  6. destruction
  7. highest
  8. lips
  9. Lips
  10. JSImport

Suppresses

  1. 品牌
  2. ensible
  3. 孩子
  4. elter
  5. 名字
  6. 填补
  7. iej
  8. rieved
  9. fname
  10. ahun