Neuron Explorer

Qwen3-0.6B

line-break tokens inside wrapped text

A neuron in layer 8 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 mostly fires on the hard line break in text that was copied with fixed-width line wrapping, where a sentence is split across lines in the middle. The steering results show no clear link to this, so the effect is weak and hard to read.

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

    The Treaty Oak has also played an important and romantic part in the history of the Lone Star State. One

  2. #2 peak

    As India dissented over⏎the US actions in Vietnam, the US offered technology as well. This⏎technology added another element that made the situation in India much⏎different and more difficult to assess than that under the British⏎Mercantilism

  3. #3 peak

    nitrogen-fixing bacteria⏎and the roots of the plants in which they live, and are examining⏎how synthetic chemicals might interfere with these signals. Disrupting⏎these 'signals of life' could have important ecosystem impacts.⏎developing fetus is

  4. #4 peak

    that: ethnically, they are the descendants of the ancient inhabitants⏎of Mesopotamia; culturally, they are the heirs of their Aramaic language⏎To be accurate from the start, I must add this clarification:⏎1) the

  5. #5 peak

    seven at the most. Atop that, the immigrants usually collected into⏎a few identifiable regions.⏎Consider the new states largest single immigrant group. Oklahomas⏎first decennial federal census, taken in 1910, calculated that

  6. #6 peak

    and⏎the occasion will be celebrated as a sign of the boy's passage into maturity⏎and promising future. A significant event in the culture and history of South⏎India is known as the Dhoti Ceremony. This is a momentous day

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 is the value of the expression 1000000000000000000000000000000000000000

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

    It’s amazing to keep in touch with everybody and see some amazing work being produced.⏎We hope

  2. #2 peak

    culture. As a whole, these examples illustrate the three kinds of discourse at which rhetoric is aimed.⏎Oxford Scholarship Online requires a subscription or purchase to access the full text of books within the service. Public users can however freely search the site and

  3. #3 peak

    the many countries whose citizens have used SOHO to chase comets.⏎Almost all of SOHO's comets are discovered using images from its Large Angle and Spectrometric Coronagraph (LASCO) instrument. LASCO is used to observe the

  4. #4 peak

    when elements cannot be saved, field biologists help bring back species and their ecosystems (restoration biology).⏎Chicago Scholarship Online requires a subscription or purchase to access the full text of

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. L2:3065 weakly suppresses
  2. L5:398 weakly excites
  3. L1:1857 weakly suppresses
  4. L7:1130 weakly suppresses
  5. L0:2871 weakly suppresses
  6. L1:1973 weakly suppresses
  7. L4:2969 weakly suppresses
  8. L2:2242 weakly suppresses
  9. L4:1307 weakly suppresses
  10. L2:111 weakly suppresses
  11. L3:2529 weakly excites
  12. L3:1392 weakly suppresses
  13. L4:245 weakly excites
  14. L7:110 weakly suppresses
  15. L6:2644 weakly excites
  16. L7:62 weakly excites
  17. L1:2278 weakly excites
  18. L2:1997 weakly suppresses
  19. L3:1948 weakly excites
  20. L4:1220 weakly suppresses
  21. L3:2820 weakly suppresses
  22. L6:2972 weakly suppresses
  23. L5:1978 weakly suppresses
  24. L7:92 weakly suppresses

Feeds later layers

  1. L9:1377 excites
  2. L9:218 excites
  3. L9:475 excites
  4. L10:923 excites
  5. L9:169 suppresses
  6. L10:710 suppresses
  7. L12:978 suppresses
  8. L19:1303 weakly excites
  9. L11:684 weakly excites
  10. L11:1686 weakly excites
  11. L9:1459 weakly excites
  12. L11:1719 weakly excites dash starting a bullet item
  13. L10:196 weakly excites
  14. L11:1233 weakly suppresses
  15. L9:134 weakly excites
  16. L11:1391 weakly excites
  17. L10:27 weakly excites
  18. L9:125 weakly excites
  19. L13:1304 weakly suppresses
  20. L24:1088 weakly suppresses
  21. L12:702 weakly excites end of text token after prose
  22. L27:2219 weakly suppresses
  23. L26:2775 weakly suppresses
  24. L21:660 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 -0.98 and -0.886 values between -0.78 and -0.689 values between -0.68 and -0.5846 values between -0.58 and -0.48131 values between -0.48 and -0.38309 values between -0.38 and -0.27558 values between -0.27 and -0.17850 values between -0.17 and -0.071095 values between -0.07 and 0.03495 values between 0.03 and 0.13216 values between 0.13 and 0.2397 values between 0.23 and 0.3348 values between 0.33 and 0.4316 values between 0.43 and 0.536 values between 0.53 and 0.6310 values between 0.63 and 0.734 values between 0.73 and 0.833 values between 0.83 and 0.931 values between 0.93 and 1.032 values between 1.03 and 1.131 values between 1.13 and 1.241 values between 1.24 and 1.341 values between 1.74 and 1.841 values between 2.04 and 2.141 values between 2.94 and 3.05offnegativepositiveoften

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. aro
  2. createQuery
  3. requently
  4. nossa
  5. ullan
  6. Morm
  7. itsu
  8. []>(
  9. Harding
  10. 我们可以

Suppresses

  1. loaded
  2. Skip
  3. 研讨会
  4. arden
  5. ��
  6. ícia
  7. phans
  8. cribe
  9. bre
  10. PointSize