Neuron Explorer

Qwen3-0.6B

question mark or line break after an FAQ heading

A neuron in layer 12 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

Fires on the question mark or line break that ends a question heading, such as "What is embalming?", just before the answer starts. Turning it up gave text that was often repetitive, so the steering does not clearly show this role.

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

    second edition of the book is produced under the sponsorship of the Linguistic Society of America and the US National Museum of Language.<|endoftext|>What is Past Perfect Tense⏎The past perfect tense describes an action completed before a certain moment in the past as

  2. #2 peak

    . Their primary advantages were the use of Iron over bronze for weapons and tools and the use of Heavy Chariots in battle.⏎Are the Hittites Greek?⏎One of a group of Indo-European speaking peoples intrusive to Anatolia, the H

  3. #3 peak

    ?⏎We looked online to see if there were any tours but couldn’t find any. Hope they offer them in the future. That hole<|endoftext|>What is Cloud Computing?⏎In simple terms, cloud computing is a technology used for storing and accessing data,

  4. #4 peak

    commonly referred to as a last name because it is usually placed at the end of a person’s full name, after any given names.⏎What exactly is a surname?⏎Also called last name . the name that a person has in common with other

  5. #5 peak

    , it is essential that Business Leaders know the AI and ML trends and can keep up with the pace and change of the business environment.⏎What are AI and ML?⏎Artificial Intelligence can be defined as a field of development of

  6. #6 peak

    ', 'loganberry'] print "You are an" , random.choice (verb) , random.choice (fruit)⏎Any Tutorial Videos?<|endoftext|>What is globalisation?⏎Globalisation is the process by which the world is becoming increasingly interconnected as a result

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 number of the number of the number of the number of the number of the number of the number of the number of the number of the number of the number of the number of the number of the number of the number of the number of

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

    two different numbers cant have the same cube). Therefore, by Definition 1, is one-to-one....⏎View Full Document⏎- Spring '09<|endoftext|>part of speech: noun⏎definition 1

  2. #2 peak

    provide a proxy)⏎|Through your library|⏎References found in this work BETA⏎No references found.⏎Citations of this work BETA⏎No citations found.⏎Similar books and articles⏎Piero Antuono & Jan Beyer (1

  3. #3 peak

    , Guatemala, Florida, and France between 1789 and 1809.2 Much remains to be known about the St. [End Page 37] Dominguan diaspora in France and in the circum-Car

  4. #4 peak

    is to cover the fundamental concepts of computer graphics, and the concentration is on graphics in three dimensions.⏎Home page url⏎Download or read it online for free here:⏎by Fränk Zimmer - Wolke Verlagsges.⏎Pd (

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. L10:259 excites
  2. L10:330 excites
  3. L10:84 suppresses
  4. L6:216 weakly excites
  5. L9:25 weakly excites
  6. L9:366 weakly excites
  7. L8:304 weakly suppresses
  8. L10:157 weakly suppresses end of document and web boilerplate
  9. L10:33 weakly suppresses
  10. L8:579 weakly excites
  11. L1:1813 weakly suppresses
  12. L8:2 weakly excites
  13. L5:21 weakly excites
  14. L7:884 weakly excites
  15. L11:775 weakly excites
  16. L8:88 weakly excites line breaks inside wrapped sentences
  17. L10:631 weakly excites
  18. L10:2985 weakly suppresses
  19. L6:0 weakly excites
  20. L4:125 weakly suppresses
  21. L9:568 weakly suppresses email address endings in examples
  22. L0:1617 weakly excites
  23. L9:1714 weakly excites
  24. L3:3 weakly suppresses

Feeds later layers

  1. L14:57 suppresses
  2. L13:2078 suppresses
  3. L14:304 suppresses
  4. L15:985 weakly excites line breaks inside wrapped text
  5. L14:1428 weakly excites
  6. L14:1321 weakly suppresses
  7. L14:3020 weakly suppresses
  8. L17:10 weakly suppresses
  9. L14:1082 weakly excites
  10. L13:2792 weakly suppresses
  11. L16:22 weakly suppresses
  12. L16:247 weakly suppresses
  13. L14:712 weakly excites
  14. L14:345 weakly suppresses
  15. L22:406 weakly suppresses
  16. L14:387 weakly excites
  17. L18:227 weakly suppresses
  18. L13:1572 weakly excites
  19. L15:227 weakly suppresses
  20. L15:615 weakly suppresses
  21. L15:550 weakly excites
  22. L15:1114 weakly suppresses
  23. L13:1343 weakly excites
  24. L14:166 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.

7 values between -0.96 and -0.7835 values between -0.78 and -0.61314 values between -0.61 and -0.431160 values between -0.43 and -0.251458 values between -0.25 and -0.07605 values between -0.07 and 0.11106 values between 0.11 and 0.2963 values between 0.29 and 0.4744 values between 0.47 and 0.6430 values between 0.64 and 0.8221 values between 0.82 and 1.0012 values between 1.00 and 1.1817 values between 1.18 and 1.369 values between 1.36 and 1.544 values between 1.54 and 1.722 values between 1.72 and 1.897 values between 1.89 and 2.072 values between 2.07 and 2.251 values between 2.25 and 2.432 values between 2.43 and 2.613 values between 2.61 and 2.791 values between 2.97 and 3.142 values between 3.32 and 3.501 values between 3.86 and 4.041 values between 4.04 and 4.221 values between 6.00 and 6.18offnegativepositiveoften

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. abouts
  2. THING
  3. 另有
  4. lining
  5. 型企业
  6. 連結
  7.     
  8. ظه
  9. evice
  10. lined

Suppresses

  1. fühl
  2. pains
  3. för
  4. phận
  5. /gin
  6. \DependencyInjection
  7. .calls
  8. ordion
  9. esis
  10. forever