Neuron Explorer

Qwen3-0.6B

word before 'that', 'to' or 'for' continuation

A neuron in layer 15 of Qwen3-0.6B. Its strongest snippets are fairly consistent: several of them peak on the same token. It is featured, so it has steering outputs.

What this neuron does

This neuron fires on a word that is about to be followed by a link word like 'that', 'to' or 'for', as in 'the energy that' or 'suitable for'. It is strongest where the sentence needs a continuation to finish a thought. Turning it up did not produce a clear theme in the model's replies.

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

    rich avifauna, characteristic of the country’s central highlands<|endoftext|>Luckily, Music Theory for Dummies shows you the fun and easy way to understanding the concepts needed to compose,

  2. #2 peak

    it to emit a great amount of radiation, mostly at visible wavelengths. The Sun size also contributes to the large amount of radiation.⏎The Sun provides all the energy that drives our climate. Global warming is a gradual increase in the average temperature of the

  3. #3 peak

    use the terms “earth”, “land”, or “ground” to refer to the same thing. Whilst you might understand soil as just the dirt that you use to pot your plants, it is so much more than that.⏎Soil is the

  4. #4 peak

    of the most influential men in his own thinking and was even accused of plagiarizing Locke while writing the Declaration of Independence.⏎In Locke we find the arguments and ideals that inspired our founding documents. For example, Locke famously argued that our dignity comes directly

  5. #5 peak

    When your cells can’t get glucose (a kind of sugar) to burn for energy, they burn fat instead. This leaves behind acids called ketones in your

  6. #6 peak

    on n-body algorithms rooted in the principles of Newtonian mechanics. The same techniques that predict planetary motions here at home in the solar system also describe the gravitational process that assembles thousands of galaxies into filaments a hundred million light-years long.⏎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: I have a 30000000000000000000000000000000000000000000

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

    ites reproduce rapidly. This allows large populations to develop and damage your tree severely before you even notice the symptoms. On top of that, oak spider mites spread throughout you and your neighbor’s property easily. Once one

  2. #2 peak

    . It is essential that you understand and comply with this legislation when purchasing an object of historical, architectural or archaeological interest. Buying with confidence will help you to comply with the legislation. Factsheet Legislation provides further information.⏎What international legislation governs the

  3. #3 peak

    the unit hydrograph.⏎Upon successful completion of this course, the student will be able to:⏎CIVL 1060 is offered as a part of the following programs:⏎If you have a comment or question about this course, please

  4. #4 peak

    about 10,180,000 square kilometres (3,930,000 sq mi), or 2% of the Earth’s surface (6.

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. L14:1547 excites
  2. L14:2163 suppresses
  3. L13:723 suppresses
  4. L12:2685 suppresses
  5. L14:2512 excites
  6. L12:2151 suppresses
  7. L13:2238 suppresses
  8. L12:2850 suppresses
  9. L14:1281 suppresses
  10. L14:576 excites
  11. L12:2694 suppresses unclear: 'purpose' and words near it
  12. L14:1821 suppresses
  13. L10:322 suppresses
  14. L12:522 suppresses
  15. L14:522 excites
  16. L10:2412 suppresses
  17. L13:2493 suppresses
  18. L12:411 excites
  19. L12:3004 excites
  20. L14:329 suppresses
  21. L11:2037 excites
  22. L13:135 suppresses document boundary end-of-text token
  23. L12:2478 suppresses
  24. L12:751 suppresses

Feeds later layers

  1. L16:1266 excites
  2. L16:347 excites
  3. L16:1540 suppresses
  4. L16:814 suppresses
  5. L18:299 suppresses
  6. L16:1008 suppresses
  7. L17:112 suppresses
  8. L16:394 excites
  9. L19:522 suppresses
  10. L17:1567 suppresses
  11. L16:316 excites
  12. L16:351 suppresses
  13. L16:2289 suppresses 'or' linking two alternatives
  14. L18:360 suppresses
  15. L17:168 excites
  16. L20:289 weakly excites
  17. L16:288 weakly excites
  18. L16:1229 weakly suppresses
  19. L18:160 weakly excites
  20. L16:1350 weakly suppresses
  21. L16:1026 weakly suppresses
  22. L19:2151 weakly suppresses
  23. L17:2764 weakly excites
  24. L16:291 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 -3.35 and -3.142 values between -2.92 and -2.711 values between -2.71 and -2.505 values between -2.50 and -2.2910 values between -2.29 and -2.0815 values between -2.08 and -1.8718 values between -1.87 and -1.6536 values between -1.65 and -1.4464 values between -1.44 and -1.23109 values between -1.23 and -1.02181 values between -1.02 and -0.81289 values between -0.81 and -0.59450 values between -0.59 and -0.38651 values between -0.38 and -0.17747 values between -0.17 and 0.04469 values between 0.04 and 0.25311 values between 0.25 and 0.47171 values between 0.47 and 0.68108 values between 0.68 and 0.8981 values between 0.89 and 1.1057 values between 1.10 and 1.3139 values between 1.31 and 1.5224 values between 1.52 and 1.7413 values between 1.74 and 1.9518 values between 1.95 and 2.1610 values between 2.16 and 2.374 values between 2.37 and 2.585 values between 2.58 and 2.808 values between 2.80 and 3.016 values between 3.01 and 3.221 values between 3.22 and 3.431 values between 3.43 and 3.642 values between 3.85 and 4.071 values between 4.91 and 5.13offnegativepositiveoften

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. Cycl
  2. eus
  3. bp
  4. mount
  5. actic
  6. ANCH
  7. arges
  8. cycl
  9. caves
  10. ROP

Suppresses

  1. \Id
  2. estureRecognizer
  3. (feed
  4. LinkId
  5. .writeValue
  6. setSearch
  7. newName
  8. ieties
  9. 娣
  10. ->{_