Neuron Explorer

Qwen3-0.6B

decimal point inside numbers

A neuron in layer 15 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 lights up on the dot in a decimal number, like the dot in 18.5 or 4.0. Turning it up made the model's writing drift into looser, more rambling sentences about random topics.

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

    are a rich source of vitamins, minerals and fiber that support optimal health.⏎A 100-gram serving of tamales contains about 18.

  2. #2 peak

    : pages Publisher: W. The Death and. Register a free business account Paperback: pages Publisher: W. Choose the book you like when you register 4.⏎Discussion questions for ‘The Death and Life of the Great Lakes’⏎Skip to

  3. #3 peak

    depending on the problem.⏎Examples of Calculations with Gravity⏎Example 1: A ball is thrown up into the air at an initial velocity of 56.3m/s. Determine its velocity after 4.52s have

  4. #4 peak

    and humor throughout the book. Make sure to find them all!⏎Where Does Art Come From? by Leah McCurdy is licensed under a Creative Commons Attribution 4.0 International License, except where otherwise noted.⏎History of art<|endoftext|>Spread over the

  5. #5 peak

    and red flag.⏎Copyright (c) 2022 Proceedings Of International Conference On Communication Science⏎This work is licensed under a Creative Commons Attribution 4.0 International License.<|endoftext|>How do you explain how babies are made to an

  6. #6 peak

    A&M College⏎Rights© 2021 Prairie View A & M University⏎This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.⏎Date of Digitization⏎John B Coleman Library⏎City

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 possible impact of the social media and the current environment. of the society and the natural environment and the focus on today's environmental and social impact of the and the the... ... 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

    of the conflict between two worlds, East and West, which determined the course of history.⏎Proving History is available online: https://2500years.culture.gov.gr/en.<|endoftext|>"Tell all the Truth"⏎Tell all the Truth

  2. #2 peak

    2021 bibliography, you will find 386 pages of resources to learn more about Caddo history and culture, https://scholarworks.sfasu.edu/cgi/viewcontent.cgi?article=4012&context

  3. #3 peak

    http://projects.fnb.nl⏎A great deal⏎of work relating to the presentation of Braille

  4. #4 peak

    . You can export a word list with a word frequency count to import into other visualization tools. http://www.antlab.sci.waseda.ac.jp/software.html⏎Voyant Tools: Voyant Tools

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. L2:9 strongly suppresses sentence-final periods vs dots in code
  2. L8:143 strongly suppresses
  3. L4:7 strongly excites
  4. L0:36 strongly excites
  5. L5:168 suppresses
  6. L5:1304 excites
  7. L4:295 excites
  8. L5:155 suppresses
  9. L10:321 suppresses
  10. L14:422 excites periods after initials and abbreviations
  11. L5:454 suppresses
  12. L2:239 suppresses
  13. L1:130 suppresses
  14. L10:156 suppresses
  15. L0:258 excites
  16. L11:132 weakly excites period ending an opening topic sentence
  17. L2:504 weakly suppresses
  18. L2:1417 weakly suppresses
  19. L1:38 weakly suppresses
  20. L4:2532 weakly excites surname pieces split across tokens
  21. L10:15 weakly suppresses
  22. L0:670 weakly suppresses
  23. L1:1256 weakly excites
  24. L2:1283 weakly suppresses

Feeds later layers

  1. L17:642 strongly excites
  2. L18:713 strongly excites
  3. L16:432 strongly excites period after month or initial
  4. L17:2084 suppresses
  5. L17:1218 suppresses
  6. L19:831 excites
  7. L16:844 excites
  8. L18:1437 excites
  9. L17:347 excites full stops ending informative sentences
  10. L17:1772 excites
  11. L19:2497 excites
  12. L17:1155 excites opening parenthesis after a sentence
  13. L18:490 excites
  14. L19:1398 suppresses
  15. L16:718 excites
  16. L19:969 excites
  17. L17:2001 excites
  18. L20:505 excites
  19. L17:1105 excites
  20. L19:902 suppresses
  21. L20:2301 excites
  22. L18:1409 excites
  23. L16:1433 excites
  24. L23:509 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.

6 values between -1.23 and -0.9985 values between -0.99 and -0.75316 values between -0.75 and -0.51935 values between -0.51 and -0.271902 values between -0.27 and -0.02491 values between -0.02 and 0.2228 values between 0.22 and 0.469 values between 0.46 and 0.7010 values between 0.70 and 0.944 values between 0.94 and 1.186 values between 1.18 and 1.4210 values between 1.42 and 1.668 values between 1.66 and 1.9012 values between 1.90 and 2.1410 values between 2.14 and 2.389 values between 2.38 and 2.629 values between 2.62 and 2.868 values between 2.86 and 3.108 values between 3.10 and 3.355 values between 3.35 and 3.597 values between 3.59 and 3.831 values between 3.83 and 4.075 values between 4.07 and 4.316 values between 4.31 and 4.554 values between 4.55 and 4.795 values between 4.79 and 5.032 values between 5.03 and 5.272 values between 5.27 and 5.511 values between 5.51 and 5.752 values between 7.20 and 7.441 values between 7.68 and 7.921 values between 8.16 and 8.40offnegativepositiveoften

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. ernet
  2. lette
  3. ussions
  4. lettes
  5. plan
  6. Plan
  7. 령
  8. illé
  9. plier
  10. ors

Suppresses

  1. target
  2. challenge
  3. curacy
  4. loom
  5. parse
  6. target
  7. focus
  8. -output
  9. prim
  10. (Parser