Neuron Explorer

Qwen3-0.6B

'as' after seen, regarded, described

A neuron in layer 1 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 reacts to the word 'as' when it follows a verb like 'seen', 'regarded' or 'described', as in 'seen as a problem'. Turning it up made the model write in a more general, wandering way, with no clear link to that word.

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

    Juan Jesus Campos and six other people were killed at the Guadalajara, Mexico, airport in what was described as a shootout involving drug gangs.<|endoftext|>Osteogenesis imperfecta (OI), also

  2. #2 peak

    stability of Australia’s electricity grid are unlikely to be made before then, large-scale electrical storage could very well have a significant role to play.⏎In what he hailed as “a

  3. #3 peak

    We have to learn how to ask questions. Make sure we leave with the understanding of what we're being told."⏎FR=EE is tackling what's been seen as a longstanding concern.⏎In 1999, the Office of Minority Health

  4. #4 peak

    he had another calling. He writes the following:⏎<|endoftext|>American music history is considered as one which has undergone tremendous transformation. In other words, what is being seen as the American music today is totally different and not close to what used to be American

  5. #5 peak

    left Germantown for Virginia because of their dissatisfaction with Lutheranis, whic was non-existent in the Colony of Pennsylvania. Matthais's name is sometimes seen as Castler. He was granted a patent of 300 acres dated September

  6. #6 peak

    . He has said:⏎“Allah does not love those who sleep too much, and do not work”.<|endoftext|>ANN ARBOR, Michigan -- Cancer is often seen as a disease of genetic changes. But one type of childhood brain tumor has stym

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 with my family, and I have been working on with my family. I have been working on with my family. I have been working on with my family. I have been working on with my family. 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

    cacao biomass – from living cacao seeds, tree, or pod tissue – and grows from them, under carefully controlled conditions, an intermediate substance (called callus) that forms the basis of the production of the cacao analogue. At present

  2. #2 peak

    can be programmed.⏎Our model no. is 884 with a serial no. of 00139⏎Kindly Donated by Callum Wilson.⏎With thanks to http://www.dvq.com/oldcomp/m

  3. #3 peak

    acid) in 1934. The use of IAA led to some success with tomato root culture (1934), and with tobacco callus development (19

  4. #4 peak

    cent.; British Mus.; see Holy Island). Carolingian minuscule script and its splendid and complex derivative, known as Gothic, were the principal calligraphic styles from the 9th to the 14th cent.⏎The

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. L0:1224 suppresses
  2. L0:1301 suppresses
  3. L0:178 suppresses
  4. L0:1001 weakly suppresses
  5. L0:1868 weakly suppresses
  6. L0:1049 weakly suppresses
  7. L0:450 weakly suppresses
  8. L0:1414 weakly excites
  9. L0:829 weakly excites
  10. L0:1503 weakly suppresses
  11. L0:558 weakly suppresses inside all-caps abbreviations
  12. L0:1475 weakly excites
  13. L0:33 weakly excites
  14. L0:1266 weakly excites
  15. L0:1241 weakly suppresses
  16. L0:636 weakly suppresses
  17. L0:1663 weakly excites
  18. L0:1374 weakly excites
  19. L0:2109 weakly excites
  20. L0:2367 weakly suppresses
  21. L0:2798 weakly suppresses
  22. L0:1109 weakly suppresses
  23. L0:69 weakly suppresses
  24. L0:953 weakly excites

Feeds later layers

  1. L21:2790 weakly excites
  2. L7:2532 weakly excites
  3. L6:1417 weakly suppresses
  4. L10:2825 weakly suppresses
  5. L20:1807 weakly suppresses
  6. L14:2650 weakly excites
  7. L9:1065 weakly excites
  8. L20:1718 weakly suppresses
  9. L10:417 weakly excites
  10. L10:2582 weakly suppresses
  11. L3:2808 weakly suppresses
  12. L11:2842 weakly suppresses
  13. L13:2392 weakly excites
  14. L15:1485 weakly suppresses
  15. L12:2134 weakly excites
  16. L14:1172 weakly excites
  17. L11:2623 weakly suppresses
  18. L20:2684 weakly suppresses
  19. L27:2357 weakly suppresses
  20. L12:720 weakly excites
  21. L10:1241 weakly excites
  22. L20:2303 weakly suppresses
  23. L6:2714 weakly suppresses
  24. L11:3058 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 -2.01 and -1.781 values between -1.78 and -1.561 values between -1.33 and -1.111 values between -1.11 and -0.893 values between -0.89 and -0.669 values between -0.66 and -0.44138 values between -0.44 and -0.212292 values between -0.21 and 0.011405 values between 0.01 and 0.2439 values between 0.24 and 0.467 values between 0.46 and 0.681 values between 0.68 and 0.914 values between 1.36 and 1.582 values between 1.58 and 1.801 values between 2.70 and 2.922 values between 4.04 and 4.271 values between 6.73 and 6.96offnegativepositiveoften

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. lier
  2. Primitive
  3. aña
  4. pit
  5. -post
  6. conquer
  7. chu
  8. -pro
  9. azz
  10. "["

Suppresses

  1. 设计方案
  2. ManagerInterface
  3. ץ
  4. 我爸
  5. .objects
  6. タイミング
  7. backgroundImage
  8. amodel
  9. ationale
  10. DialogContent