Neuron Explorer

Qwen3-0.6B

'in' after a motion verb

A neuron in layer 16 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 'in' when it follows a verb of movement, such as 'going in', 'flew in', 'heaved in' or 'flows in'. Turned down, the model drifted into number talk. Turned up, it wrote unrelated sentences, so the steering did not clearly confirm the pattern.

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

    worked, underfed horses. This action caused many streetcars to be stalled behind the first, now horseless, car. Bergh then stopped a car going in another direction.⏎The “man in the black hat” had been trying for many

  2. #2 peak

    Inside a sponge there is a system of openings and passages called a canal system. The sponge is nourished by chemicals that they filter out of the sea water passing in and out of these canals. Mature sponges do not move around.

  3. #3 peak

    more scarce because of human development, Dingoes are turning to domesticated animals for food, pitting themselves against the sheep farmers.⏎“Quite often they’ll go in and just kill a whole paddock full of sheep, and just leave it.

  4. #4 peak

    in the new park. It was chaos.⏎After much research, it was discovered that the absents of the mature bull elephants was the culprit. Quickly they flew in male role models and the bizarre and violent behavior of the juvenile elephants stopped completely in

  5. #5 peak

    a nice sheltered spot. It would even be safe from the occasional sudden great storms that struck without warning in that late summer season. As the men heaved in the anchor and the ship swung around to face the breeze, the master ordered several

  6. #6 peak

    Namibia, a former diamond mining town whose population steadily declined after WWI, when diamond prices crashed, and Balestrino, Italy, whose inhabitants were moved in 1953 after several earthquakes brought them to the conclusion that the area

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: the I: I A A A A A A A A A A A A A A A A A A A A A

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

    221BC). Unlike most Chinese idioms, 毋忘在莒 strangely has actually two stories of origin, which gave this idiom two different

  2. #2 peak

    exists in the mind as a sensation and appears to us in various hues” (Morton

  3. #3 peak

    法、分析法来进行研究,探讨泰国学生在学习汉语助动词中的困难及偏误情况,分析他们产生偏

  4. #4 peak

    to be an accessible topic for middle school students.19<|endoftext|>毋忘在莒(Pinyin: Wu Wang Zai Ju) is a famous Chinese idiom

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. L15:1328 suppresses
  2. L4:289 excites
  3. L9:1025 excites
  4. L14:1389 suppresses
  5. L3:493 suppresses
  6. L4:1966 suppresses
  7. L0:781 suppresses
  8. L10:262 suppresses
  9. L5:490 excites
  10. L1:1179 suppresses 'in' after 'result' (results in)
  11. L8:2438 suppresses
  12. L2:2307 excites
  13. L4:463 suppresses
  14. L8:351 suppresses 'in' after 'high' or 'dressed'
  15. L13:132 excites "of" after a noun
  16. L6:3070 suppresses
  17. L10:132 excites
  18. L4:2275 suppresses
  19. L3:1116 excites
  20. L0:1109 excites
  21. L14:1625 excites the word 'in' after a noun
  22. L6:846 weakly suppresses
  23. L4:16 weakly excites
  24. L4:1016 weakly excites

Feeds later layers

  1. L17:1434 strongly suppresses
  2. L17:2821 strongly suppresses
  3. L19:1345 strongly suppresses
  4. L18:1588 suppresses
  5. L17:165 suppresses
  6. L21:1269 suppresses "the" in "on the other hand"
  7. L27:1363 suppresses
  8. L19:1464 suppresses
  9. L25:1865 suppresses 'in' in 'keep in mind'
  10. L18:483 excites
  11. L23:1366 suppresses
  12. L17:2302 suppresses 'in' after a noun
  13. L17:2609 suppresses
  14. L20:2850 suppresses
  15. L23:2136 suppresses 'in' after a noun phrase
  16. L19:2775 suppresses
  17. L19:947 excites
  18. L21:3039 suppresses
  19. L19:1840 suppresses
  20. L24:1957 suppresses capital 'In' starting a word or name
  21. L18:2356 suppresses
  22. L19:2363 suppresses
  23. L17:2619 suppresses
  24. L20:2201 suppresses 'the' in location phrases like 'at the end'

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 -9.13 and -8.793 values between -8.46 and -8.132 values between -7.13 and -6.803 values between -6.47 and -6.142 values between -5.80 and -5.476 values between -5.47 and -5.143 values between -5.14 and -4.817 values between -4.81 and -4.478 values between -4.47 and -4.146 values between -4.14 and -3.8110 values between -3.81 and -3.4810 values between -3.48 and -3.154 values between -3.15 and -2.819 values between -2.81 and -2.4813 values between -2.48 and -2.1511 values between -2.15 and -1.8210 values between -1.82 and -1.4814 values between -1.48 and -1.1520 values between -1.15 and -0.8228 values between -0.82 and -0.4982 values between -0.49 and -0.161772 values between -0.16 and 0.181707 values between 0.18 and 0.51174 values between 0.51 and 0.841 values between 0.84 and 1.171 values between 2.17 and 2.501 values between 3.83 and 4.16offnegativepositiveoften

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. thái
  2. тель
  3. тельные
  4. ancement
  5. �单
  6. �
  7. мы
  8. �数
  9. ngại
  10. ��

Suppresses

  1. 屏幕上
  2. 一定程度
  3. 人群中
  4. 实践中
  5. {}]
  6. 范围内
  7. 活动中
  8. 不利
  9. vented
  10. 桌子上