Neuron Explorer

Qwen3-0.6B

symbols in headings and titles

A neuron in layer 11 of Qwen3-0.6B. Its strongest snippets are mixed: they peak on different tokens. It is featured, so it has steering outputs.

What this neuron does

This neuron fires on decorative symbols and punctuation that sit inside titles or headings on web pages, such as hearts, arrows, bars and plus signs. It is negative on full stops and hyphens inside citations and codes. Turning it up gave unrelated text, so the steering tells us little.

Written by Claude from this neuron's activations and steering outputs. Confidence: low. 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

    ly Jokes⏎☆ Cut and Paste⏎*Please look at the video preview above for a clear picture of the resource.⏎♥Why You’ll LOVE it!♥⏎Here is what teacher-buyers had to say about this resource!⏎♥ “

  2. #2 peak

    Magic Tricks for Scouts⏎> > > Triple Prizes < < <⏎piece of paper⏎- Write 1089 on the paper

  3. #3 peak

    [e^]k`t[~e]r*[=e]n"), n. [Cf.

  4. #4 peak

    AMA. To make a password even more secure, or applicable for services that require other characters, add special characters around it, like # JBS10AMA#.⏎Here are some

  5. #5 peak

    viflora. They have been sunk into E. plukenetii and are recognized as subsp. pl<|endoftext|>| ||Minding Your Mind || |⏎When Striving for Perfection Is a Problem⏎Last Reviewed by Faculty of

  6. #6 peak

    this article we explore 4 methods of how to conduct 2D lookups in Excel.⏎+++NEW+++NEW+++NEW+++ 2D XLOOKUP +++NEW+++NEW+++NEW+++⏎In a hurry?⏎Most common for 2

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 difference between the number of the first year of the month when the first time the spring and summer of the year when the spring and summer of the year when the first year of the year when the car and the plant of the plant the

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

    osenblatt, David M., "Music Listening as Therapy" (2015). Loma Linda University Electronic Theses, Dissertations & Projects. Paper 294.⏎Loma Linda University Electronic Theses and Dissert

  2. #2 peak

    ography, July 2, 2014. DOI: 10.1007/s13524-014-0308-7<|endoftext|>Designs and simulation results are given for two

  3. #3 peak

    the problem was solved.⏎Joan Marie Galat, Pelican ©2012, 978-1-4556-1637-4⏎In the 21st century, with GPS and

  4. #4 peak

    Discover the world of gorillas and other primates their remarkable intelligence and social behavior⏎66 pages; ISBN 9780756668082⏎, or download in⏎More Juvenile Nonfiction⏎IB

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. L9:102 excites
  2. L8:100 excites
  3. L7:284 suppresses
  4. L7:675 excites
  5. L6:582 weakly excites
  6. L8:991 weakly suppresses
  7. L9:284 weakly excites
  8. L0:2942 weakly suppresses
  9. L10:1431 weakly suppresses
  10. L9:1506 weakly excites
  11. L8:558 weakly excites
  12. L0:3054 weakly suppresses
  13. L7:969 weakly suppresses
  14. L10:312 weakly excites
  15. L2:2724 weakly suppresses
  16. L8:584 weakly excites
  17. L5:2161 weakly suppresses
  18. L2:1947 weakly suppresses
  19. L6:232 weakly excites
  20. L7:39 weakly excites
  21. L1:1340 weakly suppresses
  22. L4:1544 weakly suppresses
  23. L6:1779 weakly excites
  24. L5:313 weakly excites

Feeds later layers

  1. L13:525 excites second half of an opposite pair
  2. L13:1346 excites
  3. L15:1674 excites
  4. L14:2521 excites
  5. L17:2171 excites
  6. L17:1898 excites
  7. L19:358 suppresses
  8. L17:2056 excites
  9. L18:2276 weakly excites
  10. L14:1201 weakly excites
  11. L13:2230 weakly excites
  12. L15:1499 weakly excites
  13. L25:1420 weakly suppresses
  14. L16:1216 weakly excites
  15. L20:2145 weakly suppresses
  16. L21:1843 weakly suppresses
  17. L12:2886 weakly excites
  18. L15:1037 weakly excites
  19. L17:2486 weakly excites
  20. L19:2136 weakly suppresses
  21. L18:2982 weakly excites
  22. L20:2149 weakly excites
  23. L12:2020 weakly excites
  24. L19:2665 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.

2 values between -0.74 and -0.702 values between -0.63 and -0.601 values between -0.60 and -0.564 values between -0.56 and -0.5310 values between -0.53 and -0.4910 values between -0.49 and -0.4514 values between -0.45 and -0.4230 values between -0.42 and -0.3835 values between -0.38 and -0.3562 values between -0.35 and -0.3184 values between -0.31 and -0.28105 values between -0.28 and -0.24152 values between -0.24 and -0.21182 values between -0.21 and -0.17209 values between -0.17 and -0.14250 values between -0.14 and -0.10295 values between -0.10 and -0.06334 values between -0.06 and -0.03366 values between -0.03 and 0.01349 values between 0.01 and 0.04290 values between 0.04 and 0.08272 values between 0.08 and 0.11219 values between 0.11 and 0.15171 values between 0.15 and 0.19128 values between 0.19 and 0.2290 values between 0.22 and 0.2664 values between 0.26 and 0.2955 values between 0.29 and 0.3342 values between 0.33 and 0.3629 values between 0.36 and 0.4018 values between 0.40 and 0.437 values between 0.43 and 0.4712 values between 0.47 and 0.518 values between 0.51 and 0.542 values between 0.54 and 0.583 values between 0.58 and 0.612 values between 0.65 and 0.68offnegativepositiveoften

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. aurus
  2. оборот
  3. ODO
  4. :\/\/
  5. plutôt
  6. icate
  7. @endsection
  8. peu
  9. quare
  10. OU

Suppresses

  1. ItemImage
  2. nants
  3. ture
  4. (ierr
  5. ippers
  6. nown
  7. ใหญ
  8. fäh
  9. screens
  10. ierte