Neuron Explorer

Qwen3-0.6B

word endings in multi-token nouns

A neuron in layer 3 of Qwen3-0.6B. Its strongest snippets are fairly consistent: several of them peak on the same token. It is featured, so it has steering outputs.

What this neuron does

This neuron fires on the last piece of certain longer words, such as 'ies' in cardiomyopathies, 'y' in geography, or 'any' in Tuscany. It seems to care about word endings more than meaning, and turning it up did not produce a clear theme.

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

    uldade de Filosofia, Ciencias e Letras. Universidade de Sao Paulo, Zoologia,

  2. #2 peak

    the primary problem lies with the heart muscle cells themselves; these diseases are called cardiomyopathies. Different cardiomyopathies are caused by different cellular defects, but most of

  3. #3 peak

    rous Density, both direct-read ferrogram and particle quantifier, measures ferrous particles without the size limitation of emission spectroscopy.⏎- Analytical Ferrography looks at size, shape and color (the three most important physical characteristics of a

  4. #4 peak

    had a forest background. The traditions, laws, weapons, and the general approach to current problems were different. In this environment the influence of geopgaphy had charted a new approach. Modification, transformation of old traditions did not suffice

  5. #5 peak

    sauce of which you can be proud. Enjoy learning the art of sauce-making!⏎white sauce recipe⏎White sauce was originally invented in Italy (more specifically Tuscany) during the Renaissance era. Back then, it was known as salsa colla

  6. #6 peak

    formation at the time, it has been described as the embodiment of the Counter-Reformation. ⏎The Council issued condemnations of what it defined to be heresies committed by proponents of Protestantism, and also issued key statements and clarifications of

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 a 3D printer, and I have a 4D printer. So, I have a 5D printer, and I have a 6D printer. I have a 7D printer. So, I have

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

    to paints and insecticides. But recent studies by Hansen and others have shown that some silicone-based surfactants will remove spider mites, thrips and mealy bugs from apples and pears.⏎Hansen's surfactant studies at AR

  2. #2 peak

    enough water is to watch for signs of wilting. Leaves that are shriveled, papery or wilted are all signs of dehydration. Likewise, mealy bugs are a sign of overwatering.⏎A moisture meter is a great tool

  3. #3 peak

    areas on lengthy automobile trips.⏎They’re so common that you may even find them sitting atop telephone wires.⏎The Red-tailed Hawk’s loud, falling, raspy screech is frequently employed as a generic hawk or eagle call in media.⏎

  4. #4 peak

    lime, strawberries, apples, guava, lemons, mushrooms, cabbages, turnip, winter squash, potatoes, beets, carrots, leafy vegetables, pomegranates, pineapples, custard apple, as

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. L0:1921 suppresses
  2. L1:2801 excites
  3. L0:1301 suppresses
  4. L2:75 suppresses
  5. L2:1903 excites
  6. L2:50 weakly suppresses
  7. L2:1657 weakly excites
  8. L2:1321 weakly suppresses
  9. L0:178 weakly suppresses
  10. L0:1224 weakly suppresses
  11. L0:2210 weakly excites
  12. L2:1199 weakly excites
  13. L2:549 weakly suppresses
  14. L1:1973 weakly suppresses
  15. L1:1064 weakly suppresses
  16. L2:493 weakly suppresses
  17. L2:122 weakly excites
  18. L1:29 weakly excites
  19. L1:1857 weakly suppresses
  20. L0:2644 weakly suppresses
  21. L1:2922 weakly excites
  22. L1:491 weakly excites
  23. L2:1926 weakly suppresses
  24. L2:1710 weakly suppresses

Feeds later layers

  1. L4:428 strongly excites
  2. L5:983 suppresses
  3. L4:285 suppresses
  4. L4:157 suppresses
  5. L11:870 suppresses
  6. L6:1915 excites
  7. L6:105 suppresses
  8. L6:165 suppresses
  9. L19:792 suppresses
  10. L4:452 suppresses
  11. L25:1740 suppresses
  12. L10:486 suppresses
  13. L18:737 excites
  14. L16:543 excites
  15. L18:238 suppresses
  16. L7:391 suppresses 'of' after 'those' or 'many'
  17. L5:800 suppresses
  18. L17:2100 suppresses
  19. L4:128 weakly suppresses
  20. L7:400 weakly excites
  21. L12:1622 weakly suppresses
  22. L8:665 weakly suppresses
  23. L8:1346 weakly suppresses
  24. L19:2804 weakly excites 'the' after 'in' or 'use'

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 -9.70 and -9.391 values between -1.93 and -1.621 values between -1.31 and -1.001 values between -1.00 and -0.6911 values between -0.69 and -0.38845 values between -0.38 and -0.062982 values between -0.06 and 0.2564 values between 0.25 and 0.561 values between 0.87 and 1.181 values between 2.42 and 2.73offnegativepositiveoften

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. kke
  2. 涿
  3. ENDING
  4. less
  5. uxtap
  6. 钉
  7. NK
  8. homage
  9. rust
  10. arrants

Suppresses

  1. 苄
  2. веч
  3. .hw
  4. ๆ
  5. izers
  6. ness
  7. isers
  8. лат
  9. temps
  10. ##