Neuron Explorer

Qwen3-0.6B

'to' after a need or purpose

A neuron in layer 21 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 responds to the word 'to' when it introduces a purpose or need, as in 'depend on sea ice to live' or 'need water to drink'. Turning it up made the model repeat 'while' and similar small connecting words, so the link to that theme is weak.

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

    on Earth.⏎Why Does Melting Ice Matter?⏎This rise in temperature dramatically throws off the balance of the Antarctic ecosystem. Many organisms in Antarctica depend on sea ice to live. Plants and shrimplike crustaceans called krill grow on

  2. #2 peak

    you eat after a certain time, you are guaranteed to gain weight. Based on your body type and height, your body needs a specific number of calories in order to maintain your present weight. If you consume all of those calories close to bedtime,

  3. #3 peak

    Just a few months ago, wetlands were brimming with water, but now many are just parched expanses of cracked mud.⏎“Birds need water to drink every day, and when the wetlands dry out, they need to find

  4. #4 peak

    photosynthetic organisms. Hydrothermal vents, submarine hot springs, methane cold seeps, and whale falls rely only on chemical energy and chemosynthetic organisms to support life.⏎Fundamental Concept 5h.⏎Tides, waves and pred

  5. #5 peak

    might be cut down a few years later<|endoftext|>Hand-coloured lithograph entitled; 'Dampfmaschinen Pferd worauf man in einer stunde von Paris nach Petersburg reiten kann' ('Steam engine horse on which one can

  6. #6 peak

    much healthier.⏎So says endocrinologist Andrew Greenberg, director of the Obesity Metabolism Laboratory at Tufts University in Boston.⏎Your body needs fat cells to be healthy, but in obese individuals when fat cells get very big, those cells

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 most important and always **perce + *ce* + *se** * + *se* + *ce** *** + *se** * + *ce** * + *ce** * + *** + *se

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

    and air conditioning systems, it also means that pollutants that originate indoors are getting trapped inside.⏎“People normally<|endoftext|>Spring Dawn by Meng Haoran 春眠不觉晓 处处闻啼茑 夜来风雨声 花落

  2. #2 peak

    221BC). Unlike most Chinese idioms, 毋忘在莒 strangely has actually two stories of origin, which gave this idiom two

  3. #3 peak

    a matrix⏎Find nth maximum⏎Matlab Basics II - Create a vector with a repeated entry⏎Create sequnce 1 4 9 16 25.........⏎Basics

  4. #4 peak

    for the entire class.⏎PC: Windows XP and up⏎Mac: OS X (10.1) and up⏎Resolution: 800 x 600 or higher<|endoftext|>In response to the theme of 201

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. L15:184 excites
  2. L13:44 suppresses inside web addresses and citations
  3. L5:7 suppresses
  4. L11:223 weakly excites
  5. L10:155 weakly excites
  6. L18:2520 weakly excites
  7. L18:409 weakly excites
  8. L12:158 weakly suppresses
  9. L19:2264 weakly suppresses
  10. L18:274 weakly excites
  11. L18:2789 weakly suppresses
  12. L16:899 weakly excites
  13. L4:852 weakly suppresses
  14. L19:154 weakly suppresses
  15. L20:1638 weakly suppresses
  16. L18:178 weakly excites
  17. L10:1715 weakly suppresses
  18. L20:1060 weakly suppresses
  19. L15:610 weakly excites
  20. L2:75 weakly suppresses
  21. L20:2302 weakly excites
  22. L19:2388 weakly suppresses
  23. L20:2046 weakly excites
  24. L16:1479 weakly excites

Feeds later layers

  1. L23:1127 excites
  2. L22:1226 excites
  3. L22:2019 weakly excites
  4. L24:2708 weakly suppresses
  5. L27:170 weakly suppresses unclear: spaces in mixed-language text
  6. L27:760 weakly excites
  7. L23:2122 weakly excites
  8. L27:1358 weakly excites
  9. L23:1213 weakly excites
  10. L25:1381 weakly excites
  11. L22:1865 weakly suppresses
  12. L23:1926 weakly suppresses
  13. L23:1344 weakly suppresses
  14. L22:1316 weakly suppresses
  15. L25:672 weakly excites
  16. L26:637 weakly suppresses
  17. L22:267 weakly suppresses
  18. L24:2978 weakly excites
  19. L22:1694 weakly suppresses
  20. L23:769 weakly suppresses
  21. L23:2062 weakly excites
  22. L22:131 weakly excites
  23. L22:2937 weakly excites
  24. L23:278 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.

3 values between -1.40 and -1.224 values between -1.22 and -1.0314 values between -1.03 and -0.8561 values between -0.85 and -0.66191 values between -0.66 and -0.48460 values between -0.48 and -0.29904 values between -0.29 and -0.111293 values between -0.11 and 0.08628 values between 0.08 and 0.26242 values between 0.26 and 0.4564 values between 0.45 and 0.6321 values between 0.63 and 0.829 values between 0.82 and 1.003 values between 1.00 and 1.191 values between 1.19 and 1.372 values between 1.37 and 1.562 values between 1.56 and 1.741 values between 1.93 and 2.111 values between 2.30 and 2.481 values between 2.48 and 2.671 values between 2.85 and 3.042 values between 5.81 and 5.99offnegativepositiveoften

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. 而
  2. 前行
  3. 而出
  4. 前进
  5. 起
  6. while
  7. 成长
  8. while
  9. 出场
  10. 行事

Suppresses

  1. люб
  2. Stores
  3. ments
  4. iture
  5. $body
  6. jen
  7. igure
  8. sez
  9. Fixes
  10. propri