Neuron Explorer

Qwen3-0.6B

'it' referring back to a noun

A neuron in layer 24 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 'it' when it points back to a thing mentioned earlier, such as 'In other words, it's the ability' or 'what it means'. Turning it up made the output repeat 'is' and 'knows', so the steering does not match the trigger closely.

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

    did. 4⏎|NET © Notes||⏎1 tn In this place the verb חָזַק (hazaq) is used; it indicates that God would make Pharaoh’s will strong or firm.⏎2 tn The

  2. #2 peak

    vs Empathy: What Is the Difference?⏎What Is Empathy?⏎Empathy is the ability to put yourself in someone else’s shoes. In other words, it’s the ability to feel or understand another person. Experts suggest it’s innate in

  3. #3 peak

    they resemble the pointy nose and ears of a fox.⏎By the way, that likeness inspired another name — fokkusu feisu — which, because it’s written in katakana and includes the English words “fox” and “

  4. #4 peak

    78 degrees Celsius, even for a few seconds, is enough to kill most bacteria, viruses, or parasites. Toxins produced by bacteria<|endoftext|>Dit artikel staat op icthealth.nl. Link: https://ictandhealth.com/u

  5. #5 peak

    in a big world. Bravery is having the courage to go out every day and tackle what comes, no matter what comes. These creatively structured poems explore what it means to cultivate respect for self and others, the meaning of good qualities, and

  6. #6 peak

    19<|endoftext|>毋忘在莒(Pinyin: Wu Wang Zai Ju) is a famous Chinese idiom (成語, Pinyin: Cheng Yu). It is sort of like the Latin idiom left in the European language. Most are

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: when I start a new project, I start with the most recent recent recent recent recent recent recent recent recent recent recent recent recent recent recent recent recent recent recent... (I start writing, but I don're...) 都说都都说都都说

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

    varied, from about 5.5 to 8 inches (13<|endoftext|>by Kong Dehao⏎The search for extra-terrestrial intelligence (SETI) celebrates its 58th anniversary this year (Nature, 20

  2. #2 peak

    had twice been admitted for “exhaustion” and doctors were observing the chronically fatigued young man. Beds of wounded men lined room around him.⏎Everything fell to a disciplined hush when the six-foot, two-inch General George S

  3. #3 peak

    on account of race or color.⏎With the Civil Rights Act turning 50, the success of the struggle for equal rights for women and minorities<|endoftext|>The No Child Left Behind Act, signed into law by President Bush January 8, 2

  4. #4 peak

    been caught by the Roman authorities, but others had renounced Christianity in order to stay alive. At first no one had much choice in the matter; too few clergy were willing to make themselves known

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. L22:309 strongly suppresses
  2. L20:1423 strongly excites
  3. L19:1317 strongly suppresses
  4. L21:468 strongly suppresses
  5. L23:625 strongly excites
  6. L23:176 strongly excites
  7. L23:1210 strongly excites
  8. L21:2936 strongly excites
  9. L23:287 strongly excites 'are' after a plural subject
  10. L19:1422 strongly suppresses
  11. L20:2453 strongly suppresses blank slot before a date
  12. L19:545 suppresses
  13. L18:444 excites
  14. L21:262 suppresses
  15. L21:853 suppresses 'can' after 'and' or a noun subject
  16. L19:1743 suppresses
  17. L19:1424 suppresses
  18. L23:1832 excites
  19. L19:1139 suppresses
  20. L23:2540 excites
  21. L19:880 excites
  22. L23:320 excites
  23. L22:228 suppresses
  24. L17:1583 suppresses

Feeds later layers

  1. L25:586 strongly excites
  2. L27:321 strongly excites
  3. L26:467 strongly excites
  4. L27:461 strongly excites
  5. L25:347 strongly excites
  6. L26:1097 strongly excites
  7. L26:1844 strongly excites
  8. L25:108 strongly suppresses
  9. L26:926 strongly excites
  10. L26:72 strongly suppresses 'they' or 'we' after where or who
  11. L27:500 excites
  12. L25:104 excites
  13. L26:669 excites
  14. L26:2147 excites
  15. L26:874 suppresses
  16. L25:629 excites
  17. L27:480 excites
  18. L26:764 excites
  19. L27:1313 excites
  20. L27:969 excites
  21. L27:60 suppresses
  22. L25:364 suppresses
  23. L27:1256 excites
  24. L26:676 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 -21.78 and -20.542 values between -16.81 and -15.571 values between -15.57 and -14.331 values between -14.33 and -13.091 values between -11.84 and -10.602 values between -10.60 and -9.365 values between -9.36 and -8.129 values between -8.12 and -6.886 values between -6.88 and -5.6313 values between -5.63 and -4.3929 values between -4.39 and -3.1565 values between -3.15 and -1.91321 values between -1.91 and -0.662509 values between -0.66 and 0.58600 values between 0.58 and 1.82153 values between 1.82 and 3.0657 values between 3.06 and 4.3127 values between 4.31 and 5.5522 values between 5.55 and 6.7921 values between 6.79 and 8.0310 values between 8.03 and 9.2713 values between 9.27 and 10.5211 values between 10.52 and 11.767 values between 11.76 and 13.004 values between 13.00 and 14.244 values between 14.24 and 15.482 values between 15.48 and 16.731 values between 16.73 and 17.973 values between 17.97 and 19.212 values between 19.21 and 20.453 values between 21.70 and 22.942 values between 25.42 and 26.661 values between 26.66 and 27.91offnegativepositiveoften

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. has
  2. $is
  3. hasn
  4. knows
  5. comes
  6. includes
  7. requires
  8. goes
  9. is
  10. applies

Suppresses

  1. 都是
  2. 都要
  3. 均
  4. 都不
  5. 都无法
  6. 都
  7. 都不会
  8. 都非常
  9. ayers
  10. 都可以