Neuron Explorer

Qwen3-0.6B

opening parenthesis after a term

A neuron in layer 14 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 an opening bracket "(" that comes right after a name or term and adds a short explanation, such as an abbreviation or an alternative label. Its negative side is a different pattern: the digit 6 inside the HTTP code "406 Not Acceptable". Turning it up did not produce clearly related writing.

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

    the best way to impress your teacher with your thesis. A Guide to Paraphrasing in Research Papers (APA, AMA ... In all research papers with formatting guidelines (APA, AMA, MLA, etc.), quoted text must be accompanied by quotation marks

  2. #2 peak

    ||Classical and Quantum Gravity|⏎|State||Published - Jul 7 2003|⏎ASJC Scopus subject areas⏎- Physics and Astronomy (miscellaneous)<|endoftext|>Most computer users are familiar with the concept of spyware —

  3. #3 peak

    year produces over $350,000 more income for his/her students compared to an average teacher. On the the other hand, a poor teacher (one in the bottom quarter in terms of effectiveness) subtracts $350

  4. #4 peak

    carcinomatosis—are cancers that are frequently referred to as peritoneal cancers. Whether peritoneal cancer starts in the peritoneum or spreads from somewhere else (colon⏎Ovarian Epithelial, Fallopian Tube, and Primary

  5. #5 peak

    2K).⏎GLib (pronounced ``gee-lib'') shared libraries are a way to implement shared libraries on the Pilot that differs from the ``standard'' (SysLib) mechanism for the Palm Pilot.⏎Some features of GLib:⏎GL

  6. #6 peak

    America for all of the estimated 1.1 billion people around the world who depend on unimproved drinking water sources. With the predicted increase of precipitation variability (Marengo, 2009), it is likely that diarrhea rates

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 just read, I think, and I have to do, but I don't know, it's not possible, not forget, I can't. I don't need not, I don't can't, no, I don't

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

    the search results.⏎In the example, the client requests a resource in the French language but the server responds with the 406 Not Acceptable response because only German and English versions of the document are available.⏎

  2. #2 peak

    , a behavior that is considered ethical or socially acceptable in one community may be utterly unacceptable<|endoftext|>406 Not Acceptable⏎HTTP response status code 406 Not Acceptable is a client error that is returned by the server to indicate that

  3. #3 peak

    used because the server supplies a default representation instead. The rationale is that it assumes that the client will prefer something over nothing. However, when the 406 Not Acceptable response is returned, the message body will contain a list of available

  4. #4 peak

    VU past papers, and links to you. To join WhatsApp groups click the below links.⏎MUST JOIN VU STUDY GROUPS⏎CS610: Computer Networks⏎A network is defined as a system for connecting computers using a

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. L13:2024 suppresses
  2. L12:2240 suppresses
  3. L9:1065 suppresses
  4. L12:2171 suppresses
  5. L12:2293 excites
  6. L12:2893 suppresses
  7. L13:112 suppresses
  8. L6:990 excites
  9. L12:1443 excites opening parenthesis after a name or term
  10. L13:846 suppresses
  11. L12:2947 excites
  12. L9:1370 suppresses
  13. L12:3064 excites
  14. L6:2511 excites
  15. L13:1501 weakly suppresses
  16. L1:2560 weakly excites
  17. L12:2539 weakly suppresses
  18. L11:2400 weakly excites
  19. L10:297 weakly excites
  20. L13:517 weakly suppresses comma closing a list item
  21. L12:2100 weakly suppresses opening parenthesis in citations
  22. L11:1307 weakly excites 'words' in 'in other words'
  23. L10:362 weakly suppresses
  24. L13:431 weakly excites

Feeds later layers

  1. L15:508 suppresses
  2. L16:1322 suppresses
  3. L20:271 excites
  4. L15:2956 suppresses
  5. L16:2918 suppresses
  6. L17:208 suppresses
  7. L20:251 excites
  8. L15:404 excites
  9. L20:27 suppresses
  10. L18:675 suppresses
  11. L18:435 suppresses
  12. L16:896 suppresses
  13. L18:156 suppresses
  14. L18:176 suppresses
  15. L15:1438 suppresses
  16. L16:1113 suppresses
  17. L20:186 suppresses
  18. L16:2725 suppresses
  19. L15:396 excites
  20. L16:1267 suppresses
  21. L18:1173 suppresses
  22. L17:1022 weakly suppresses
  23. L20:1430 weakly suppresses
  24. L24:150 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 -2.14 and -2.011 values between -2.01 and -1.891 values between -1.89 and -1.762 values between -1.51 and -1.391 values between -1.27 and -1.144 values between -1.14 and -1.024 values between -1.02 and -0.897 values between -0.89 and -0.775 values between -0.77 and -0.6515 values between -0.65 and -0.5228 values between -0.52 and -0.40121 values between -0.40 and -0.27394 values between -0.27 and -0.151041 values between -0.15 and -0.031537 values between -0.03 and 0.10512 values between 0.10 and 0.22159 values between 0.22 and 0.3545 values between 0.35 and 0.4715 values between 0.47 and 0.604 values between 0.60 and 0.721 values between 0.72 and 0.841 values between 0.84 and 0.972 values between 0.97 and 1.092 values between 1.09 and 1.222 values between 1.22 and 1.341 values between 1.84 and 1.961 values between 2.21 and 2.331 values between 2.71 and 2.83offnegativepositiveoften

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. izers
  2. TabIndex
  3. económico
  4. icios
  5. isers
  6. foods
  7. izr
  8. [word
  9. ife
  10. .microsoft

Suppresses

  1. tha
  2. otic
  3. pecting
  4. "|
  5. aign
  6. artial
  7. athy
  8. athan
  9. 여
  10. orting