Neuron Explorer

Qwen3-0.6B

"as" after "known"

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 lights up on the word "as" when it comes straight after "known", as in "known as quercetin" or "known as Streptococcus". It is a clear and reliable pattern. Turning it up did not bring out this theme in the model's replies, so that test tells us nothing either way.

Written by Claude from this neuron's activations and steering outputs. Confidence: high. 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

    rich in vitamins, minerals, carbohydrates and amino acids etc. Apples are also very good source of vitamin B-Complex.⏎Apple contains a kind of antioxidant known as ‘quercetin’ which is very essential for neurological health. It reduces

  2. #2 peak

    in an enzyme known as argininosuccinate synthase which is needed for this vital process, leading

  3. #3 peak

    . The trash accumulates from all over the world and as it moves into the ocean, it becomes trapped at the center of a system of rotating ocean currents known as the

  4. #4 peak

    ) is a proto-oncogene that encodes a protein known as hepatocyte growth factor receptor (HGFR). The hepatocyte growth factor receptor protein

  5. #5 peak

    If strep has been diagnosed, a lack of improvement after taking antibiotics for 48 hours⏎Strep throat is caused by infection with a bacterium known as Streptococcus pyogenes, also called group A streptococcus

  6. #6 peak

    in the slave quarters. Sometimes the white masters would come to the slave quarters to watch the celebration. Parents would give children small, homemade tokens.⏎Another celebration known as Jonkonnu, or a Christmas masquerade, took place on the plant

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 know, I know, I know, I know, I know, I know, I know, I know, I know, I know, I know, I know, I know, I know, I know, I know

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

    're gone.⏎The title of this well-known story comes from a poem by Sara Teasdale in which nature is depicted as indifferent

  2. #2 peak

    both sides. Part A: Short Answer (1 problems, 12 points).

  3. #3 peak

    and Livia Visser-Fuchs:⏎“A ‘Most Benevolent Queen’: Elizabeth Woodville’s Reputation, her Piety and her Books”, includes a convincing refutation of common allegations

  4. #4 peak

    slave economy.⏎The name of a former kingdom at the confluence of the Niger and Benue Rivers in West Africa

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. L0:2554 weakly excites
  2. L0:1089 weakly excites
  3. L0:1224 weakly suppresses
  4. L0:178 weakly suppresses
  5. L0:1152 weakly suppresses
  6. L0:218 weakly excites
  7. L0:1960 weakly suppresses unclear: mixed medical and harsh words
  8. L0:1133 weakly suppresses
  9. L0:1301 weakly suppresses
  10. L0:932 weakly excites
  11. L0:2853 weakly excites
  12. L0:2257 weakly suppresses
  13. L0:862 weakly suppresses
  14. L0:12 weakly suppresses
  15. L0:1001 weakly suppresses
  16. L0:392 weakly suppresses
  17. L0:183 weakly suppresses
  18. L0:2986 weakly suppresses
  19. L0:976 weakly suppresses
  20. L0:711 weakly excites
  21. L0:2684 weakly excites
  22. L0:2529 weakly excites
  23. L0:1066 weakly excites
  24. L0:71 weakly excites

Feeds later layers

  1. L7:79 excites
  2. L19:1934 excites
  3. L16:1599 excites
  4. L8:2006 excites
  5. L13:838 weakly excites
  6. L20:954 weakly excites
  7. L5:680 weakly excites
  8. L9:2895 weakly excites
  9. L16:1734 weakly excites
  10. L6:1639 weakly excites
  11. L11:1051 weakly excites
  12. L18:469 weakly suppresses
  13. L18:1678 weakly excites
  14. L19:1897 weakly excites
  15. L17:1322 weakly excites
  16. L16:2697 weakly excites
  17. L17:1841 weakly excites
  18. L3:402 weakly excites
  19. L19:2068 weakly suppresses
  20. L12:636 weakly excites
  21. L10:2627 weakly excites
  22. L11:1306 weakly excites
  23. L4:1004 weakly excites
  24. L21:362 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.

2 values between -0.89 and -0.7919 values between -0.68 and -0.5838 values between -0.58 and -0.4857 values between -0.48 and -0.3866 values between -0.38 and -0.28143 values between -0.28 and -0.18572 values between -0.18 and -0.081600 values between -0.08 and 0.021023 values between 0.02 and 0.13278 values between 0.13 and 0.2371 values between 0.23 and 0.3315 values between 0.33 and 0.439 values between 0.43 and 0.534 values between 0.63 and 0.732 values between 0.73 and 0.833 values between 0.83 and 0.931 values between 0.93 and 1.041 values between 1.04 and 1.142 values between 1.14 and 1.241 values between 1.54 and 1.641 values between 3.06 and 3.16offnegativepositiveoften

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. umu
  2. onder
  3. unteer
  4. bler
  5. ousel
  6. auer
  7. дум
  8. ues
  9. igmat
  10. uet

Suppresses

  1. prere
  2. void
  3. .fd
  4. iect
  5. Dll
  6. rtl
  7. ящих
  8. retr
  9. 瘪
  10. 还原