Neuron Explorer

Qwen3-0.6B

'of' after abstract nouns like 'sense' or 'view'

A neuron in layer 2 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 responds to the word 'of' that follows nouns such as 'sense', 'view', 'need' or 'idea', as in 'sense of injustice'. Its negative side picks up words like 'inception' or 'absence' that follow 'their' or 'since'. Turning it up did not give a clear theme in the model's writing.

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

    glimpse into the academic world and its view of religion in general and Hinduism in particular. The hundred-plus pages of articles are

  2. #2 peak

    form of English was essentially a request from the customers (i.e. the Airlines), of which 80% are not native English speakers, and their need of doing correct maintenance to guarantee the aircraft availability.⏎Complex technical instructions can be misunderstood and

  3. #3 peak

    stand to see animals killed just for fun. Her sense of injustice makes her angry and through her anger she develops a special power - not diss

  4. #4 peak

    ?⏎- Do the characters/narrator reflect upon whether they exist and how they know what they know?⏎- To what extent do the following factors impact their sense of self: emotions, past

  5. #5 peak

    small as asteroids go. Because of that (and its current distance of 450 million kilometers) it's a bit faint, shining right now

  6. #6 peak

    or with a jacket when going out, the shirts relaxed basket weave also allowed them to be worn pretty much year round. Today, practically every company produces their version of the OCBD, and they’re somewhat of a collectors item, not because they

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 have been thinking about lately: I have been thinking about lately: I have been thinking about lately: I have been thinking about lately: I have been thinking about lately: I have been thinking about lately: I have been thinking about lately

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

    Europe where it formed an intricate part of military bands. Pipe and drum bands were quite popular and predate military bands with a wide variety of instruments. Since their inception, snare drums can be heard in virtually all types of European and North American

  2. #2 peak

    ment. This has now been banned, captive breeding sites for each species has been established, but the problems involved in saving these birds are immense.⏎In their absence, populations of feral dogs and rats have taken over their scavenging role

  3. #3 peak

    is subject to mechanical forces from intramolecular tension? An important component of the proposed project is technology development, which has been a hallmark of this grant since its inception in the 1980s, when we established the first quantitative in

  4. #4 peak

    take up torches and pitchforks against him on the silver screen. Never once during his rampage, however, did the monster question the honesty of his ultimate creator,

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. L0:326 weakly excites
  2. L1:993 weakly suppresses
  3. L1:433 weakly suppresses
  4. L1:2902 weakly suppresses
  5. L1:2362 weakly excites
  6. L1:2979 weakly suppresses
  7. L1:2801 weakly excites
  8. L0:1523 weakly excites
  9. L1:454 weakly excites
  10. L1:370 weakly excites
  11. L0:414 weakly suppresses blank after a dash in number ranges
  12. L0:121 weakly excites
  13. L1:1993 weakly excites
  14. L0:620 weakly suppresses
  15. L1:1340 weakly suppresses
  16. L1:2925 weakly excites
  17. L1:383 weakly excites
  18. L0:1745 weakly excites quantity word after large, high or low
  19. L1:426 weakly excites
  20. L0:1750 weakly excites
  21. L0:1878 weakly excites
  22. L1:1419 weakly excites
  23. L0:670 weakly suppresses
  24. L1:752 weakly excites

Feeds later layers

  1. L3:4 excites
  2. L7:1821 suppresses
  3. L12:575 excites
  4. L7:1624 suppresses
  5. L9:69 excites
  6. L4:81 suppresses
  7. L14:282 excites
  8. L4:510 excites
  9. L3:1015 suppresses
  10. L9:537 excites
  11. L4:72 suppresses
  12. L22:1051 suppresses
  13. L3:58 weakly excites
  14. L3:402 weakly suppresses
  15. L9:1182 weakly suppresses
  16. L5:453 weakly suppresses
  17. L9:1412 weakly suppresses the word "be" after modals and "to"
  18. L10:235 weakly excites
  19. L6:2221 weakly suppresses
  20. L6:285 weakly suppresses
  21. L5:305 weakly excites
  22. L3:471 weakly suppresses
  23. L5:804 weakly suppresses
  24. L6:101 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 -1.86 and -1.791 values between -1.59 and -1.522 values between -0.91 and -0.841 values between -0.77 and -0.701 values between -0.70 and -0.641 values between -0.64 and -0.572 values between -0.57 and -0.503 values between -0.50 and -0.435 values between -0.43 and -0.3714 values between -0.37 and -0.3025 values between -0.30 and -0.23161 values between -0.23 and -0.16402 values between -0.16 and -0.09941 values between -0.09 and -0.031357 values between -0.03 and 0.04653 values between 0.04 and 0.11212 values between 0.11 and 0.1875 values between 0.18 and 0.2525 values between 0.25 and 0.3118 values between 0.31 and 0.385 values between 0.38 and 0.451 values between 0.45 and 0.521 values between 0.58 and 0.651 values between 0.79 and 0.86offnegativepositiveoften

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. -type
  2. type
  3. -Type
  4. -lo
  5. ชน
  6. ươi
  7. -rel
  8. lo
  9. tipo
  10. -text

Suppresses

  1. ');</
  2. acie
  3. _history
  4. atIndex
  5. 灯火
  6. anale
  7. 历史上
  8. zza
  9. 髯
  10. 琬