Neuron Explorer

Qwen3-0.6B

the 'inter' word beginning

A neuron in layer 26 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 "inter" part at the start of words like interstates, interplay, interwoven and interrelated, and sometimes on a stray number. Turned up, the writing fell into loops of the word "down" or "single", which is not clearly related.

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

    of its market value. To the unprofessional eye the object might look good, but the truth lies in the details: use of the correct reversible materials; intrusive interventional methods kept to a minimum; careful documentation of the research and the work process

  2. #2 peak

    years Hartford has lost 10,000 residents every decade, and getting through and around the City by automobile is often frustrating and complicated. Two elevated interstates slice through Hartford, and its explosively growing suburbs come at the expense of

  3. #3 peak

    of Chhayavadi school of Hindi poetry which gained popularity in late 19th and early 20th centuries.⏎Kamayani depicts the interplay of human emotions, thoughts, and actions by taking mythological metaphors.

  4. #4 peak

    pier arches, consisting of ornamental trees and flowering shrubs, lay the burial grounds. The choir featured a colourful roof of red and green leafy trees interwoven with Indian laburnum flowers. Behind the ‘choir fence’

  5. #5 peak

    the validity and worth of the evidence, bias and prejudice become difficult to remove. These arguments are amplified by two interrelated aspects of 21st-century life: firstly, the “post-tr

  6. #6 peak

    a complex and inter-connected microbial neighborhood that appears to be associated with asthma, and akin to what has

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 do you do when your host hosters host themselves themselves themselves themselves themselves themselves themselves themselves themselves themselves themselves themselves themselves themselves themselves themselves themselves themselves themselves themselves themselves themselves themselves themselves themselves themselves themselves themselves themselves themselves themselves themselves themselves themselves themselves themselves themselves themselves

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

    name Kudara 1. This name in Japan refers to the Korean state of Paekche⏎百濟 2, which at the time of the Three States (Samguk 三國) was occupying the South-Western part

  2. #2 peak

    Western pubblications they are designated with: "dw." and "jw." - they are abbreviations of Polish words: "dwuwieżowy" = twin-turreted and "jednowieżowy"

  3. #3 peak

    garden bed by using a garden fork or tiller to loosen the soil to a depth of 12 to 15 inches, then mix in a 2- to 4-inch layer of compost and/or aged manure.⏎Planting

  4. #4 peak

    berg, who -- like them -- had hung around CDC Corporation near the University of Washington cadging open time on the mainframe. Wennberg, founder of the Triakis Corporation, was then an electrical engineering student at the University of Washington. In

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. L19:1752 excites
  2. L20:311 suppresses
  3. L22:2314 excites
  4. L7:520 excites
  5. L7:2731 weakly excites
  6. L16:161 weakly suppresses
  7. L2:2600 weakly suppresses
  8. L4:326 weakly suppresses
  9. L21:708 weakly suppresses
  10. L20:2616 weakly suppresses
  11. L4:776 weakly excites
  12. L0:1010 weakly suppresses
  13. L19:1042 weakly excites
  14. L17:391 weakly excites
  15. L16:860 weakly excites
  16. L17:1818 weakly excites
  17. L19:723 weakly excites
  18. L17:1594 weakly excites
  19. L1:2935 weakly suppresses
  20. L16:600 weakly suppresses
  21. L2:127 weakly excites
  22. L13:3021 weakly suppresses
  23. L21:2561 weakly excites
  24. L3:2629 weakly excites

Feeds later layers

  1. L27:2751 weakly suppresses
  2. L27:2021 weakly excites
  3. L27:698 weakly suppresses
  4. L27:704 weakly suppresses
  5. L27:1320 weakly excites
  6. L27:757 weakly suppresses
  7. L27:2199 weakly suppresses
  8. L27:860 weakly suppresses
  9. L27:1684 weakly suppresses
  10. L27:1800 weakly suppresses
  11. L27:3024 weakly suppresses
  12. L27:2526 weakly suppresses
  13. L27:2141 weakly excites
  14. L27:1749 weakly suppresses hyphen joining word parts
  15. L27:1460 weakly suppresses
  16. L27:2178 weakly suppresses
  17. L27:1958 weakly excites
  18. L27:1554 weakly suppresses
  19. L27:2047 weakly excites
  20. L27:1515 weakly suppresses
  21. L27:930 weakly suppresses
  22. L27:2142 weakly excites
  23. L27:2901 weakly excites
  24. L27:2152 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 -39.13 and -36.801 values between -27.51 and -25.181 values between -25.18 and -22.864 values between -13.57 and -11.243 values between -11.24 and -8.924 values between -8.92 and -6.608 values between -6.60 and -4.2718 values between -4.27 and -1.953538 values between -1.95 and 0.37245 values between 0.37 and 2.7029 values between 2.70 and 5.0217 values between 5.02 and 7.3413 values between 7.34 and 9.6711 values between 9.67 and 11.999 values between 11.99 and 14.312 values between 14.31 and 16.642 values between 16.64 and 18.961 values between 21.28 and 23.611 values between 51.49 and 53.81offnegativepositiveoften

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. toa
  2. icit
  3. aina
  4. 一夜
  5. we
  6. PasswordField
  7. itm
  8. ếp
  9. boxed
  10. WAY

Suppresses

  1. yourselves
  2. themselves
  3. 所在
  4. &_
  5. しい
  6. noun
  7. النقد
  8. ьте
  9. 中关
  10. urile