Neuron Explorer

Qwen3-0.6B

'of' before a measured value

A neuron in layer 20 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 word 'of' when it comes right before a measurement or quantity, like 'a depth of 12 m' or 'efficiency of 4-6%'. It switches off on words like 'a' or 'low' that lead into a number. Turning it up made the model start writing lists of numbers and percentages.

Written by Claude from this neuron's activations and steering outputs. Confidence: high. 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

    options for artificial teeth replacements, the natural teeth will always be the preferred teeth of professionals.⏎By<|endoftext|>Demonstration: This rough method can produce a difference of mass of at least 8 g . With enough apparatus

  2. #2 peak

    by the authors.<|endoftext|>Image: Common Seadragon, Phyllopteryx taeniolatus⏎A male Common Seadragon at a depth of 12 m at 'two poles', off Bateau Bay, New South

  3. #3 peak

    — A thin-film, silicon photovoltaic cell having no crystalline structure. Manufactured by depositing layers of doped silicon on a substrate. Efficiency of panels is generally 4-6% and performance does degrade rapidly - as much

  4. #4 peak

    to have an asymmetry in the normal population, having a leftward bias in right-handed individuals, people who possess absolute pitch have an increased leftward asymmetry of the planum temporale. This

  5. #5 peak

    -aircraft coverage, and even sniping ability. Its combination of exceptional mobility – being light enough to be carried by one man – and high rate of fire (of up to 900 rounds per minute) was unmatched. It entered service

  6. #6 peak

    2.2% and their mean agreed with that obtained by the Kjeldahl method. By use of the prescribed dilution of the sample, detection limits of 0.1% can be achieved for all three cations. ??

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 future of a health system for the future. It is a health system that will have no single, clear, public health sector, and that will have a more efficient and diverse, and more transparent, and more sustainable, and more efficient

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

    0 to $102,202 per year with median annual wages estimated at $49,490. Hourly rates range from a low of $13.50 to a high of $49.0

  2. #2 peak

    of injections performed in the United States, estimated from Medicare procedure codes, increased from fewer than 3000 per year in 1999 to an estimated 6.5 million in 2016. This number continues

  3. #3 peak

    50 years; further research shows turtle dove numbers have plummeted by 80 per cent since 1995, hedgehog numbers are down by a third since 20

  4. #4 peak

    700 among adults ages 35 and 44, more than 5,600 among adults 45 to 54, more than 3,800 among adults 55 to 64

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. L19:248 strongly excites
  2. L19:1211 strongly excites
  3. L19:1222 strongly suppresses
  4. L19:3047 suppresses
  5. L18:2183 suppresses
  6. L10:868 suppresses
  7. L19:1921 excites
  8. L17:1372 suppresses
  9. L19:1454 excites
  10. L17:2230 suppresses
  11. L18:390 excites
  12. L17:2083 excites
  13. L14:641 excites
  14. L6:1149 suppresses
  15. L18:2150 excites
  16. L7:1524 suppresses
  17. L16:3042 suppresses
  18. L14:471 excites
  19. L18:2384 suppresses
  20. L19:803 excites
  21. L6:1008 excites 'of' after a measured quantity
  22. L14:1793 suppresses
  23. L6:756 suppresses
  24. L18:291 suppresses

Feeds later layers

  1. L21:676 strongly excites
  2. L21:175 strongly excites
  3. L24:722 strongly excites 'about' or 'approximately' before a number
  4. L26:667 strongly excites
  5. L25:2902 strongly excites
  6. L25:253 strongly excites
  7. L24:587 excites
  8. L25:46 suppresses
  9. L23:198 excites
  10. L27:287 excites
  11. L24:126 excites
  12. L26:476 excites
  13. L21:336 excites
  14. L27:2172 suppresses
  15. L26:423 suppresses
  16. L22:1146 excites blank before a time or count
  17. L26:172 excites
  18. L24:162 suppresses
  19. L21:543 excites
  20. L25:2908 excites
  21. L27:113 suppresses
  22. L25:1212 excites
  23. L25:1136 excites
  24. L27:163 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 -18.89 and -18.091 values between -17.30 and -16.501 values between -12.52 and -11.721 values between -11.72 and -10.921 values between -9.33 and -8.531 values between -7.74 and -6.943 values between -6.14 and -5.353 values between -5.35 and -4.555 values between -4.55 and -3.755 values between -3.75 and -2.969 values between -2.96 and -2.1622 values between -2.16 and -1.36388 values between -1.36 and -0.573066 values between -0.57 and 0.23356 values between 0.23 and 1.0315 values between 1.03 and 1.8213 values between 1.82 and 2.625 values between 2.62 and 3.422 values between 3.42 and 4.212 values between 4.21 and 5.012 values between 5.01 and 5.811 values between 5.81 and 6.602 values between 6.60 and 7.401 values between 8.20 and 8.991 values between 8.99 and 9.791 values between 12.18 and 12.98offnegativepositiveoften

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. fifty
  2. 五百
  3. 五千
  4. thirty
  5. 六十
  6. 三十
  7. forty
  8. 三千
  9. 约
  10. 四十

Suppresses

  1. eve
  2. enza
  3. ']}}</
  4. endphp
  5. güneş
  6. AVE
  7. IMP
  8. council
  9. eenth
  10. ilename