Neuron Explorer

Qwen3-0.6B

opening parenthesis after a name

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 fires on the open bracket that comes right after a name or term, as in a full name followed by its short form or dates, like 'Point Calimere Wildlife Sanctuary (PCWS)'. Turning it up made the model write lists of numbers or bracketed options, which is only loosely related.

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

    <|endoftext|>Jacques-Louis David (1748-1825) is generally considered the leading figure

  2. #2 peak

    access to information and communication technologies (ICTs) can mitigate many issues in SIDS and help to diversify their

  3. #3 peak

    was aimed to investigate the patterns of liana-tree interactions in tropical dry evergreen forest of Point Calimere Wildlife Sanctuary (PCWS), south- east India. This study also analyzes the impact of various

  4. #4 peak

    temporoparietal junction (RTPJ), an area involved in mental state reasoning, is necessary for making

  5. #5 peak

    Silla (57 BC[note 1] – 935 AD) (

  6. #6 peak

    system for diagnosis of HIV infection (Weiss et al., 1985). As a result of this

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: by and as, and any and in the but and and and so and but and so and as and and and but and any and and and and and the and and and so and and and with and and and so and and and in

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

    how these relate to what we know today.⏎In association with STFC Rutherford Appleton Laboratory⏎If your child has additional needs, please get in touch on [email protected] to discuss how we can best support you.⏎For information on our

  2. #2 peak

    anything with sugar in it⏎- anything with wheat, corn in it⏎- dog food⏎- cat food (mots of it)⏎What your Ferret<|endoftext|>The 2018

  3. #3 peak

    essay, we will delve into one of Mary Oliver’s poems, analyzing its themes, literary devices, and the impact it has on readers.⏎The Poem: [Insert Poem Title]⏎[Insert the full text of the chosen Mary Oliver poem

  4. #4 peak

    other fluids.⏎- Avoid exercising for at least 8 hours.⏎- 24 hours without alcohol.⏎- Have not smoked for 4 hours before the measurement (including e-cigarette with nicotine).⏎The measurement of basic metabolism in the diet office

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. L18:980 suppresses opening parenthesis after a noun phrase
  2. L17:304 excites opening parenthesis after a claim
  3. L16:1513 excites opening parenthesis before a numbered list
  4. L17:238 excites
  5. L16:264 suppresses opening parenthesis in citations and asides
  6. L10:66 suppresses opening parenthesis after a citation or term
  7. L16:869 excites
  8. L16:710 suppresses
  9. L11:50 excites
  10. L17:1708 excites
  11. L18:1273 suppresses
  12. L18:682 weakly excites
  13. L10:1041 weakly suppresses opening parenthesis before phone or code
  14. L1:2526 weakly excites
  15. L14:90 weakly suppresses
  16. L5:330 weakly excites opening parenthesis after a noun
  17. L9:339 weakly excites
  18. L4:2309 weakly excites
  19. L16:183 weakly suppresses
  20. L17:2160 weakly suppresses
  21. L16:170 weakly suppresses
  22. L8:464 weakly excites
  23. L4:1799 weakly excites
  24. L17:35 weakly suppresses

Feeds later layers

  1. L20:36 suppresses
  2. L24:148 suppresses
  3. L24:35 excites
  4. L21:200 weakly excites
  5. L27:193 weakly suppresses
  6. L26:211 weakly excites
  7. L23:528 weakly excites
  8. L24:328 weakly suppresses
  9. L20:813 weakly suppresses
  10. L27:173 weakly excites
  11. L24:1690 weakly excites
  12. L20:1629 weakly suppresses
  13. L20:66 weakly suppresses
  14. L20:348 weakly suppresses
  15. L25:1270 weakly excites
  16. L22:267 weakly excites
  17. L21:478 weakly suppresses
  18. L25:476 weakly suppresses comma before 'but' in contrasts
  19. L27:19 weakly excites
  20. L22:160 weakly excites
  21. L27:78 weakly excites
  22. L27:289 weakly suppresses
  23. L27:139 weakly suppresses
  24. L21:524 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.

4 values between -1.83 and -1.26364 values between -1.26 and -0.682545 values between -0.68 and -0.11842 values between -0.11 and 0.4762 values between 0.47 and 1.0531 values between 1.05 and 1.6214 values between 1.62 and 2.209 values between 2.20 and 2.774 values between 2.77 and 3.354 values between 3.35 and 3.927 values between 3.92 and 4.505 values between 4.50 and 5.081 values between 5.08 and 5.651 values between 5.65 and 6.231 values between 6.23 and 6.801 values between 6.80 and 7.382 values between 7.95 and 8.531 values between 8.53 and 9.102 values between 9.10 and 9.681 values between 10.83 and 11.411 values between 12.56 and 13.132 values between 13.13 and 13.711 values between 14.86 and 15.431 values between 16.01 and 16.581 values between 17.74 and 18.311 values between 20.61 and 21.19offnegativepositiveoften

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. /
  2. .
  3. (
  4. /
  5. (
  6. ·
  7. ·
  8. q
  9. done

Suppresses

  1. 但由于
  2. ignant
  3. theless
  4. 但这
  5. uste
  6. }?
  7. 的同时
  8. ,**
  9. 但他
  10. mientras