Neuron Explorer

Qwen3-0.6B

the zero in round numbers like 100

A neuron in layer 5 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 zero inside numbers like 100, 1000 or 300, and on the word "seventy." It is about the size of a round number being written out. Turning it up made the model talk about numbers and counting, which is 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

    74 Illingworth accompanied George Armstrong Custer’s expedition into the Black Hills, and he captured over seventy images. Illingworth was taken to court for embezzling government property

  2. #2 peak

    Valley called TransitCamp. Inspired by a similar event in Toronto, the idea was to brainstorm what you might do with transit-agency data. Nearly 100 people came. One guy was looking to build a Web site that combined an online

  3. #3 peak

    It's likely to get lost. Your character will appear wh<|endoftext|>The 45th annual PDK/Gallup Poll collected data from over 1000 Americans ages 18+ regarding the approach to improving the

  4. #4 peak

    stability and serenity. Other significant plants on the property are especially notable for their maturity and precise placement. Some rhododendrons are over 100 years old, their gnarled trunks as attractive as their flowers. Tr

  5. #5 peak

    -25 percent higher than houses with no trees.⏎Trees generate jobs and contribute raw materials for buildings, newspapers, books and more than 15,000 other forest products. Trees are renewable, biodegradable and recyclable.

  6. #6 peak

    regulating your intake and staying healthy, both physically and mentally.<|endoftext|>Tech moves fast! Stay ahead of the curve with Techopedia!⏎Join nearly 200,000 subscribers who receive actionable tech insights from Techopedia.⏎A

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: 1/1, 2/2, 3/4, 5/5. 5/5? Well, I need 50000 to be 10000? Let me try that

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

    reformer” led to his⏎appointment as head of Scots Church, Collins Street (1875-83) in Melbourne.⏎Yet, “almost from the hour of his arrival”, controversy was never far away⏎and his essays,

  2. #2 peak

    , which heightens the intensity of this sound installation. The volume also oscillates—some parts are loud, others whir and buzz at levels that are almost inaudible, then it becomes tinkly and jolting before cycling back to

  3. #3 peak

    and Ukraine, approximately in the 10th century. The Cossacks were known for their military

  4. #4 peak

    excess weight, more than it is considered healthy for their height and age.⏎People with this disorder may go to great lengths to diet, almost to the point of starving or exercise even if they are underweight.⏎It’s all

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. L4:2624 suppresses
  2. L3:830 weakly excites
  3. L4:546 weakly suppresses
  4. L4:1905 weakly suppresses
  5. L4:2951 weakly excites
  6. L2:2968 weakly excites
  7. L1:1810 weakly excites
  8. L2:1002 weakly excites
  9. L1:2607 weakly suppresses
  10. L3:2111 weakly excites
  11. L3:632 weakly excites
  12. L1:1047 weakly excites
  13. L2:1874 weakly suppresses
  14. L4:3049 weakly suppresses
  15. L3:1086 weakly suppresses
  16. L3:2492 weakly suppresses
  17. L3:453 weakly suppresses
  18. L1:2797 weakly suppresses
  19. L0:1067 weakly suppresses
  20. L2:2789 weakly excites
  21. L1:2278 weakly excites
  22. L3:2703 weakly suppresses
  23. L0:593 weakly excites
  24. L3:3048 weakly suppresses 'called' after 'so'

Feeds later layers

  1. L10:1636 excites
  2. L9:620 excites
  3. L7:534 excites 'much' or 'many' after 'not' or 'little'
  4. L7:755 weakly excites
  5. L15:2484 weakly excites
  6. L22:1736 weakly excites
  7. L6:1037 weakly excites
  8. L16:1273 weakly excites
  9. L15:2350 weakly excites
  10. L11:2495 weakly excites
  11. L9:781 weakly excites
  12. L8:391 weakly excites
  13. L7:245 weakly excites 'even' in 'and even' lists
  14. L11:2797 weakly excites
  15. L18:1309 weakly excites
  16. L7:561 weakly suppresses 'all' or 'every' before a group
  17. L13:1083 weakly suppresses
  18. L8:2933 weakly suppresses
  19. L10:1063 weakly excites
  20. L26:1200 weakly excites
  21. L22:1056 weakly suppresses
  22. L6:1877 weakly excites
  23. L25:1731 weakly excites
  24. L20:4 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 -1.67 and -1.4111 values between -0.89 and -0.6357 values between -0.63 and -0.37424 values between -0.37 and -0.113124 values between -0.11 and 0.15224 values between 0.15 and 0.4125 values between 0.41 and 0.678 values between 0.67 and 0.937 values between 0.93 and 1.192 values between 1.19 and 1.451 values between 1.45 and 1.713 values between 1.71 and 1.974 values between 1.97 and 2.232 values between 2.23 and 2.493 values between 2.49 and 2.751 values between 2.75 and 3.012 values between 3.01 and 3.271 values between 3.53 and 3.791 values between 3.79 and 4.052 values between 4.31 and 4.571 values between 5.34 and 5.601 values between 6.38 and 6.641 values between 6.90 and 7.161 values between 7.94 and 8.201 values between 8.46 and 8.72offnegativepositiveoften

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. EPROM
  2. ander
  3. inct
  4. adero
  5. agem
  6. raph
  7. ��
  8. .dw
  9. dpi
  10. !"⏎

Suppresses

  1. 格
  2. 不对
  3. inen
  4. ais
  5. 使者
  6. 推进
  7. galement
  8. 第一节
  9. orney
  10. hungry