Neuron Explorer

Qwen3-0.6B

the :// in web addresses

A neuron in layer 19 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 colon and slashes in the middle of a web address, right after words like http or https. Turning it up did not produce anything web-related in the sample outputs, so its effect on writing is unclear.

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

    [change | change source]⏎- "Chomsky is citation champ". MIT News Office. 1992-04-15. http://web.mit.edu/newsoffice/1992/citation-0

  2. #2 peak

    1: Download Drupal⏎First download the Drupal using following command:⏎# wget http://ftp.drupal.org/files/projects/drupal-7.33.zip⏎Step

  3. #3 peak

    is available for this purpose (e.g., US Census data at https://www.census.gov/) and can be of tremendous help in targeting various ethnic groups

  4. #4 peak

    structure = function.⏎(Image from: https://commons.wikimedia.org/w/index.php?curid=12679

  5. #5 peak

    for efficient DNA cloning and recombineering in E. coli". PLoS ONE 8 (2): e57075. . . http://www.plosone.org/article/info%3Adoi%2F10

  6. #6 peak

    lenses, similar to radiotelescopes arrays.⏎The limitation is givven by the Index of Refraction. For more information you can explore this: https://en.wikipedia.org/wiki/Refractive_index⏎Regarding the image, it is

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: how to actually * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

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

    , rule, minor head injury⏎Ann Afr Surg. 2018;15(2):57-61 DOI:http://dx.doi.org/10.4314/aas.v15i2

  2. #2 peak

    Red List of Threatened Species 2016: e.T6794A21952571. https://dx.doi.org/10.2305/IUCN.UK.2

  3. #3 peak

    /2002, WGBH Media Library & Archives, accessed March 29, 2017, http://openvault.wgbh.org/catalog/V_4E4EE3EAD24043

  4. #4 peak

    Recommendation on<|endoftext|>- OurWorld 2.0 - http://ourworld.unu.edu/en -⏎What does Cancún offer the climate generation?⏎Posted By Huw

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. L10:2362 suppresses
  2. L10:814 suppresses
  3. L5:2579 excites
  4. L15:1962 suppresses
  5. L18:2292 suppresses
  6. L10:613 excites
  7. L18:569 excites
  8. L6:2232 weakly excites
  9. L0:1536 weakly excites
  10. L15:2745 weakly excites
  11. L14:2306 weakly suppresses word after a motion or location verb
  12. L4:26 weakly excites
  13. L5:50 weakly excites
  14. L11:2017 weakly excites
  15. L3:456 weakly suppresses
  16. L9:1040 weakly excites
  17. L9:656 weakly suppresses
  18. L6:18 weakly excites
  19. L14:1505 weakly excites
  20. L18:1451 weakly excites
  21. L4:2427 weakly suppresses
  22. L13:1372 weakly excites number before an age or week count
  23. L4:1796 weakly suppresses
  24. L5:120 weakly suppresses

Feeds later layers

  1. L24:228 excites
  2. L21:1855 excites
  3. L20:2623 excites
  4. L20:536 excites
  5. L21:1033 excites '://' in web addresses
  6. L25:656 excites
  7. L22:406 weakly excites
  8. L23:372 weakly excites
  9. L26:1066 weakly excites
  10. L27:580 weakly excites
  11. L25:294 weakly excites
  12. L27:362 weakly excites
  13. L20:1111 weakly suppresses
  14. L26:593 weakly excites
  15. L23:1308 weakly excites
  16. L27:879 weakly excites
  17. L21:389 weakly excites
  18. L26:316 weakly excites
  19. L24:1427 weakly excites
  20. L25:1135 weakly excites
  21. L20:327 weakly excites
  22. L23:2581 weakly excites
  23. L20:2748 weakly suppresses
  24. L21:1825 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 -9.70 and -9.071 values between -7.18 and -6.55155 values between -0.89 and -0.263713 values between -0.26 and 0.3835 values between 0.38 and 1.011 values between 2.27 and 2.901 values between 14.23 and 14.861 values between 14.86 and 15.49offnegativepositiveoften

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. www
  2. www
  3. .example
  4. �
  5. ền
  6. -static
  7. .stack
  8. ritis
  9. .bias
  10. example

Suppresses

  1. ertino
  2. 还是要
  3. 其实
  4. 依旧
  5. 其實
  6. 依然
  7. 属实
  8. 实际
  9. LEncoder
  10. 仍