Neuron Explorer

Qwen3-0.6B

'in' after motion verbs

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 fires on the word "in" when it follows verbs of moving or pushing, like "move in packs" or "pushing in all directions". Turning it up did not give a clear, consistent change in the model's output.

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

    ales are usually smaller. They are generally about 5-7 feet long from nose to tail (the tail is usually about 2 feet long) and move in packs of up to 15 members! To learn more about gray wolves,

  2. #2 peak

    tall cone such as Liberty Cap increases in height, the flow of water from the top may eventually stop, because there is

  3. #3 peak

    production of plastic materials has steadily increased in recent years. Because of their versatile and resilient nature, plastics play a critical role

  4. #4 peak

    forward direction - that is, it thrusts the airplane forward.⏎Jet propulsion is similar to the release of an inflated balloon. The pressure inside the balloon is pushing in all directions. It is also "jetting" out from the mouth of the

  5. #5 peak

    international, dry bulk goods use bulk ships. Liquid bulk moves in ocean-going tankers. General cargoes move in container ships or break bulk ships

  6. #6 peak

    0 minutes – it’s essentially moving in a straight line. The reason the path is all bendy is because Swift itself

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 working on this. I have been working on this. I have been working on this. I have been working on this. I have been working on this. I have been working on this. I have been working on this

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

    “white” turkeys found in North America are actually Broad-breasted Whites, which are a cross between White Holland

  2. #2 peak

    . This can also cause dehydration and vitamin and mineral deficiencies. Treatment may require hospitalization.⏎HG is a more severe form of nausea and vomiting of pregnancy (NVP), also called morning sickness. Morning sickness affects anywhere between 50% to

  3. #3 peak

    their ability to meet and worship, and has helped people in prisons, jails, mental institutions, and state-run nursing homes preserve their religious freedom.⏎The values embodied in RLUIPA are universal ideals. Department of Justice attorneys have provided technical assistance on

  4. #4 peak

    of inaccuracies and misfires.⏎Truth, when hyped, loses credibility. There have been attempts, of late, to eulogise ancient theories contained in ayurvedic classics as ultra-modern ‘quantum physiology’1.

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. L0:38 weakly suppresses
  2. L0:781 weakly suppresses
  3. L1:2413 weakly excites
  4. L1:1110 weakly excites
  5. L0:880 weakly excites
  6. L0:2641 weakly suppresses
  7. L0:1311 weakly suppresses
  8. L0:2162 weakly suppresses
  9. L0:1316 weakly suppresses
  10. L0:1405 weakly excites
  11. L1:1724 weakly excites
  12. L1:2904 weakly suppresses
  13. L0:204 weakly suppresses
  14. L1:1131 weakly suppresses
  15. L1:2859 weakly excites
  16. L1:2722 weakly suppresses
  17. L0:2374 weakly excites
  18. L1:1346 weakly excites
  19. L0:624 weakly suppresses
  20. L0:1224 weakly suppresses
  21. L1:2042 weakly excites
  22. L0:178 weakly suppresses
  23. L1:1 weakly suppresses
  24. L0:2951 weakly excites

Feeds later layers

  1. L3:2778 weakly suppresses
  2. L13:2527 weakly suppresses
  3. L14:1939 weakly suppresses
  4. L5:2204 weakly excites
  5. L13:1213 weakly suppresses
  6. L7:2562 weakly suppresses
  7. L5:2849 weakly excites
  8. L13:383 weakly suppresses
  9. L6:1997 weakly excites
  10. L13:1921 weakly excites
  11. L16:2903 weakly suppresses
  12. L9:363 weakly excites
  13. L10:2908 weakly excites
  14. L12:544 weakly suppresses
  15. L3:184 weakly suppresses
  16. L16:607 weakly suppresses
  17. L12:1870 weakly suppresses
  18. L8:932 weakly suppresses
  19. L19:2494 weakly suppresses
  20. L11:2176 weakly suppresses
  21. L11:123 weakly suppresses
  22. L24:2849 weakly suppresses
  23. L9:2521 weakly suppresses
  24. L15:336 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 -1.24 and -1.141 values between -1.14 and -1.031 values between -0.73 and -0.622 values between -0.62 and -0.523 values between -0.52 and -0.4218 values between -0.42 and -0.3192 values between -0.31 and -0.21457 values between -0.21 and -0.111409 values between -0.11 and -0.001407 values between -0.00 and 0.10388 values between 0.10 and 0.2074 values between 0.20 and 0.3112 values between 0.31 and 0.415 values between 0.41 and 0.517 values between 0.51 and 0.614 values between 0.61 and 0.722 values between 0.72 and 0.824 values between 0.82 and 0.926 values between 0.92 and 1.032 values between 1.03 and 1.131 values between 1.13 and 1.231 values between 1.23 and 1.341 values between 1.34 and 1.441 values between 1.44 and 1.541 values between 1.54 and 1.651 values between 1.65 and 1.751 values between 1.75 and 1.851 values between 1.85 and 1.952 values between 2.06 and 2.161 values between 2.16 and 2.262 values between 2.78 and 2.88offnegativepositiveoften

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. '~
  2. imeType
  3. pecies
  4. order
  5. leness
  6. ưỡ
  7. olars
  8. ited
  9. :type
  10. trình

Suppresses

  1. gage
  2. unted
  3. stra
  4. ANCH
  5. doubted
  6. ATEST
  7. amento
  8. .stem
  9. _CAST
  10. aten