Neuron Explorer

Qwen3-0.6B

blank spot before a run-on number

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 the empty gap just before a number that has been attached to the end of a sentence or header, such as a footnote number, a comment count, or a time after a date. Turning it up made the model produce repetitive lists and labels, which doesn't clearly match this pattern.

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

    a good footnote and give you a place to start if you develop constant and persistent joint pain, but they are not predictive and certainly not diagnostic.⏎| Related: 9 Everyday Surprises When RA Enters Your Life |⏎Many of these risk factors

  2. #2 peak

    games [The Daily Journal]<|endoftext|>Should You Ask Your Children to Apologize?By Craig Smith | September 23, 2016 | 0 comments⏎When kids say sorry, are they learning a lesson or just par

  3. #3 peak

    evidence indicating that living systems may be designed.⏎The Kansas Citizens for Science has recently urged 1 you to take the former path, to teach only naturalism. 2 They have asked you to reject the Science Education Standards adopted by the Kansas State

  4. #4 peak

    , turn by turn directions, UI, User interface, video chat, Virtual reality, Virtual reality glasses⏎January 18, 2012 • 11:07 am

  5. #5 peak

    5, for example, a separate team developed a tiny, $0.20 temperature sensor that didn’t require a battery or constant source of wired electricity. 3D printing isn't new either, and already, researchers can print out human

  6. #6 peak

    Struggle-For-Power-In-The-994323.html⏎"The Struggle for Power in the Ussr" StudyMode.com. 05 2012. 05 2012

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:,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,

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

    Arabian lava field from the International Space Station:⏎Harrat Khaybar, Saudi Arabia is featured in this image photographed by an Expedition 16 crewmember on the International Space Station. The western half of the Arabian

  2. #2 peak

    Exercise 6.5⏎In NCERT Solutions for Class 10 Maths Chapter 6 Exercise 6.5, we have discussed the

  3. #3 peak

    body feels no pain.⏎Let your child know that these cremated remains are placed in a⏎container called an urn and returned to the family. If cremation⏎has already taken place and the container picked up, you may want⏎to show

  4. #4 peak

    rules, where⏎are they all sitting? Can you find all the possible ways?⏎What is the best way to shunt these carriages so that each train⏎can continue its journey?⏎If you split the square into these

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. L9:7 excites
  2. L8:30 excites blank slot right before a number
  3. L6:0 excites
  4. L1:1 excites
  5. L10:33 suppresses
  6. L12:5 excites
  7. L3:3 suppresses
  8. L7:961 excites
  9. L3:691 weakly excites
  10. L1:108 weakly suppresses
  11. L8:4 weakly excites
  12. L0:2098 weakly excites
  13. L8:16 weakly suppresses
  14. L8:10 weakly excites
  15. L0:1247 weakly suppresses
  16. L0:324 weakly suppresses
  17. L8:14 weakly excites
  18. L6:1263 weakly suppresses
  19. L8:2 weakly excites
  20. L3:264 weakly suppresses
  21. L6:4 weakly suppresses
  22. L7:63 weakly excites
  23. L6:16 weakly excites
  24. L7:122 weakly excites

Feeds later layers

  1. L22:372 excites empty spot before a number or "3D"
  2. L18:106 excites
  3. L16:1939 excites line break mid-sentence in wrapped text
  4. L19:781 excites
  5. L21:173 excites
  6. L17:359 excites
  7. L25:118 excites
  8. L16:822 excites
  9. L15:123 excites
  10. L23:77 excites the space right before a number
  11. L17:23 excites
  12. L17:1787 excites
  13. L18:429 excites
  14. L20:1067 excites
  15. L27:598 excites
  16. L24:130 excites line break at the end of a line
  17. L22:1744 excites
  18. L21:189 excites document break token mid-sentence
  19. L19:581 excites blank slot before a number
  20. L23:35 excites
  21. L15:122 excites
  22. L15:1870 excites
  23. L19:182 suppresses
  24. L18:2153 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.

9 values between -0.93 and -0.8140 values between -0.81 and -0.69188 values between -0.69 and -0.57437 values between -0.57 and -0.45843 values between -0.45 and -0.331057 values between -0.33 and -0.21756 values between -0.21 and -0.08262 values between -0.08 and 0.0460 values between 0.04 and 0.1643 values between 0.16 and 0.2827 values between 0.28 and 0.4029 values between 0.40 and 0.5223 values between 0.52 and 0.6428 values between 0.64 and 0.7619 values between 0.76 and 0.8817 values between 0.88 and 1.0017 values between 1.00 and 1.1215 values between 1.12 and 1.246 values between 1.24 and 1.365 values between 1.36 and 1.484 values between 1.48 and 1.615 values between 1.61 and 1.732 values between 1.85 and 1.976 values between 1.97 and 2.091 values between 2.09 and 2.213 values between 2.21 and 2.331 values between 2.45 and 2.573 values between 2.81 and 2.932 values between 3.78 and 3.90offnegativepositiveoften

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. undry
  2. 本报讯
  3. issued
  4. occan
  5. .addComponent
  6. gay
  7. 让用户
  8. 霸王
  9. ITHER
  10. 곱

Suppresses

  1. leshoot
  2. Wer
  3. entimes
  4. ification
  5. нтер
  6. atoon
  7. rze
  8. irling
  9. 为主
  10. areth