Neuron Explorer

Qwen3-0.6B

commas in keyword lists

A neuron in layer 3 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 commas that separate items in a list of keywords or tags, especially near the end of the list. It goes negative on the middle of some longer words. Turning it up or down did not produce a clear, consistent change in the text.

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

    central china, chinese government, conservation efforts, early 1970, giant panda, giant pandas, hsing hsing, ling ling, mainland china, mei xiang, mother and baby, mountain forests, national zoo, pandas

  2. #2 peak

    General technology, Google, Media, Services, Augmented reality, eyeglasses, feature, featured, glasses, Google, Google Plus, Google Project Gass, google virtual reality glasses, New York Times, Project Glass, technology, turn by

  3. #3 peak

    asymmetric radial transects with fixed orientation, but a workable solution for polygonal transects remains elusive.⏎Keywords: Boundary overlap, Inclusion region, Reflection method, Transect configuration⏎- We recommend that you also print this page and attach it

  4. #4 peak

    , Hau believes future practical applications include improved communications technology, television displays and night-vision devices.⏎Tags: Lene Vestergaard Hau, physicist, spped of light.<|endoftext|>The Real

  5. #5 peak

    Gaming, Services, Software, 3 million dollar grant, Associate Professor, Augmented reality, Bill & Melinda Gates Foundation, Common Core, Common Core standards, Education, education arcade, education program, Eric Klopfer, Filament Games

  6. #6 peak

    and reduce inequalities in household decision-making.⏎keywords malaria, treatment-seeking behavior, inequality in household decision-making, Ghana, child mortality, chloroquine, home treatment, pre-packaged antimalarial drugs, prevention⏎Malaria, a mosquito-b

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 am a good person. I have a good home. I am a good friend to the Earth. I am a good planet. I am a good... and I am a good... and I am a good... It's the first

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

    elage. It was under the umbrella of American power that today even the smallest European state can enjoy its national identity and independence in unprecsented security. Domestically America in the 20th century gradually implemented a series of progressiveand

  2. #2 peak

    voltage drop waveforms are in-phase with each other.⏎Given that the power supply frequency is 60 Hz (household power frequency in the United States), calculate impedances for all components and determine all

  3. #3 peak

    bacteria Yersinia pestis, and can be contracted by humans and household animals if proper precautions are not taken,” officials from Jefferson County Public Health (J

  4. #4 peak

    , which was published in 1979, the year she died.⏎‘Impossible to understate Rhys’ importance’⏎Jean Rhys is a household name across the Caribbean due in large part to the fact that “Wide Sarg

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. L2:733 weakly excites
  2. L2:2558 weakly suppresses
  3. L2:1036 weakly excites
  4. L2:2346 weakly excites
  5. L1:583 weakly suppresses
  6. L2:1657 weakly excites
  7. L2:41 weakly suppresses
  8. L2:2784 weakly suppresses
  9. L2:1494 weakly suppresses
  10. L1:649 weakly excites
  11. L2:1710 weakly suppresses
  12. L0:1046 weakly suppresses
  13. L1:2463 weakly excites
  14. L2:75 weakly suppresses
  15. L1:2160 weakly suppresses
  16. L0:895 weakly excites
  17. L2:929 weakly suppresses
  18. L1:3038 weakly excites
  19. L2:122 weakly excites
  20. L1:997 weakly suppresses
  21. L2:2149 weakly suppresses
  22. L0:1512 weakly suppresses
  23. L1:2605 weakly suppresses
  24. L1:597 weakly excites

Feeds later layers

  1. L12:552 weakly excites
  2. L18:1935 weakly excites
  3. L7:2713 weakly suppresses
  4. L16:1722 weakly excites
  5. L17:2618 weakly suppresses
  6. L6:938 weakly excites
  7. L16:589 weakly suppresses
  8. L10:2863 weakly excites
  9. L7:511 weakly suppresses
  10. L19:1315 weakly suppresses
  11. L10:1150 weakly suppresses
  12. L4:2137 weakly suppresses
  13. L12:471 weakly suppresses
  14. L4:831 weakly suppresses
  15. L14:1020 weakly excites
  16. L9:775 weakly excites
  17. L7:1031 weakly suppresses
  18. L10:865 weakly suppresses
  19. L6:2934 weakly suppresses
  20. L23:1120 weakly excites
  21. L4:1126 weakly suppresses
  22. L15:1853 weakly excites
  23. L11:2383 weakly suppresses
  24. L11:2416 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 -6.32 and -5.961 values between -3.37 and -3.002 values between -1.53 and -1.161 values between -1.16 and -0.7918 values between -0.79 and -0.42863 values between -0.42 and -0.052972 values between -0.05 and 0.3237 values between 0.32 and 0.695 values between 0.69 and 1.061 values between 1.06 and 1.431 values between 1.43 and 1.804 values between 1.80 and 2.161 values between 2.90 and 3.271 values between 8.07 and 8.44offnegativepositiveoften

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. extent
  2. ológ
  3. deadline
  4. CD
  5. PTS
  6. 长大了
  7. 我爱你
  8. درك
  9. grade
  10. 幅度

Suppresses

  1. etwork
  2. pomi
  3. culator
  4. atab
  5. �
  6. idian
  7. iphers
  8. lycer
  9. pee
  10. dition