Neuron Explorer

Qwen3-0.6B

'://' in web addresses

A neuron in layer 21 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 the colon and slashes that come right after 'http' or 'https' in a web address. Turning it up or down did not make the model write about links or web addresses, so the steering output does not confirm the pattern.

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

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

  2. #2 peak

    413967-1-768x146-1-300x57.png 0 0 admin https://

  3. #3 peak

    ://my.fsf.org/donate/working-together/octave.⏎DocumentationUser manual included; User FAQ included; Printed and online user manual available from http://www.network-theory.co.uk/octave/manual/⏎- PARI GP⏎-

  4. #4 peak

    to supports all kinds of powerful, dynamic applications and services. For more information on this great topic, please consult:⏎- Sun's Java Server Pages home (http://java.sun.com/products/jsp/)⏎- Westy Rockwell's article on "

  5. #5 peak

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

  6. #6 peak

    restriction of intentional use in many industrial sectors as detergents, fertilizers, construction and oil&gas materials. The documents are available at this link (https://echa.europa.eu/it/registry-of-restriction-intentions/-/dis

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: the.m. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

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

    there, I get an exception, which is life as usual for a Python programmer.⏎if not server_response['successful']:⏎etag = server_response['file_handle'].info().getheader('ETag')⏎xml_document = server_response['xml_document']

  2. #2 peak

    achieved with the following code:⏎TYPE List = REF RECORD next: List; value: INTEGER; END;⏎VAR x := NEW(List, value:=1); BEGIN x.next :=

  3. #3 peak

    syntax in configuration files is case-sensitive.⏎The machine configuration file, Machine.config, contains settings that apply to an entire computer. This file is located in the %runtime

  4. #4 peak

    , instead of writing Users.Current.User.Name, you can just write Users.Current.Name.⏎I’ll grant you that this doesn’t seem much for the User entity

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. L16:1696 excites
  2. L18:2709 suppresses
  3. L19:1049 suppresses
  4. L20:2113 excites
  5. L17:2932 excites
  6. L19:913 excites
  7. L19:1766 suppresses
  8. L12:1039 excites
  9. L16:2648 suppresses
  10. L19:1400 excites
  11. L20:1340 suppresses
  12. L11:2255 excites
  13. L19:3009 excites
  14. L14:1248 suppresses
  15. L9:2559 excites
  16. L19:2234 excites the :// in web addresses
  17. L15:2615 excites
  18. L11:1047 weakly excites pieces of software and library names
  19. L19:274 weakly suppresses
  20. L8:2364 weakly excites
  21. L10:845 weakly excites
  22. L18:1565 weakly suppresses
  23. L15:653 weakly excites
  24. L14:2515 weakly suppresses

Feeds later layers

  1. L24:228 strongly suppresses
  2. L26:593 strongly suppresses
  3. L25:294 suppresses
  4. L23:372 suppresses
  5. L27:362 suppresses
  6. L26:174 suppresses
  7. L27:580 suppresses
  8. L26:943 suppresses
  9. L26:1169 suppresses
  10. L27:1328 suppresses
  11. L25:387 suppresses
  12. L27:816 suppresses
  13. L25:412 suppresses
  14. L27:351 suppresses
  15. L27:875 suppresses
  16. L26:316 suppresses
  17. L25:1135 suppresses
  18. L27:370 suppresses
  19. L27:1240 suppresses
  20. L25:656 suppresses
  21. L27:2971 weakly suppresses
  22. L22:1829 weakly suppresses
  23. L26:167 weakly suppresses
  24. L26:653 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 -2.19 and -2.111 values between -1.54 and -1.461 values between -1.06 and -0.981 values between -0.98 and -0.901 values between -0.90 and -0.813 values between -0.81 and -0.734 values between -0.73 and -0.656 values between -0.65 and -0.576 values between -0.57 and -0.4924 values between -0.49 and -0.4148 values between -0.41 and -0.3375 values between -0.33 and -0.25164 values between -0.25 and -0.17342 values between -0.17 and -0.09683 values between -0.09 and -0.01926 values between -0.01 and 0.08582 values between 0.08 and 0.16400 values between 0.16 and 0.24251 values between 0.24 and 0.32164 values between 0.32 and 0.4097 values between 0.40 and 0.4854 values between 0.48 and 0.5644 values between 0.56 and 0.6413 values between 0.64 and 0.727 values between 0.72 and 0.804 values between 0.80 and 0.893 values between 0.89 and 0.972 values between 0.97 and 1.05offnegativepositiveoften

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. わり
  2. 倒在
  3. ностей
  4. 同一个
  5. Meadows
  6. 人々
  7. 켈
  8. 許多
  9. 否則
  10. 好几个

Suppresses

  1. .s
  2. .S
  3. .b
  4. .Mon
  5. .F
  6. .f
  7. .L
  8. .M
  9. .m
  10. .v