Neuron Explorer

Qwen3-0.6B

'and' joining two full clauses

A neuron in layer 19 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 lights up on the word 'and' when it links two complete statements, such as 'this browser will no longer be supported and you may experience issues'. When it was turned up, the model fell into repeated filler words like 'is is is', and when turned down it rambled on with long, wandering sentences.

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

    not support the requirements of modern web applications and pose significant security risks. After May 1, 2017, this browser will no longer be supported and you may experience issues or not be able to access the site at all. We

  2. #2 peak

    the recipient at another branch of the same bank. These items are not cleared between institutions and therefore are not represented in statistics of payments exchanged between financial institutions in the clearings

  3. #3 peak

    , other sources of individual variation, or environmental conditions.⏎You have requested a machine translation of selected content from our databases. This functionality is provided solely for your convenience and is in no way intended to replace human translation. Neither BioOne nor the owners

  4. #4 peak

    a distance of about 20 km from Una town and is one of the most visited pilgrimage sites in the region. The temple is dedicated to Lord Shiva and is surrounded by beautiful natural surroundings, making it an ideal place for meditation and spiritual

  5. #5 peak

    fourth-most populous county. Its county seat and largest city is Ogden, the home of Weber State University. The county was formed in 1850 and named after the Weber River, which in turn was named for John Henry Weber (

  6. #6 peak

    fly. Despite your best efforts, sometimes the way you care for the gardenia may encourage whitefly infestations.⏎Whiteflies resemble tiny grains of white rice and tend to congregate on the underside of the gardenia’s foliage. They thrive

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: "If I, with the help of a new tool, that which I previously unable to do, and which I previously unable to do with a new tool, but with a new one, which I now able to do, which I now

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

    a common food source for many Native American tribes including but not limited to the Iroquois, Algonquin, Ojibwa

  2. #2 peak

    Athenae (see Athens).⏎(a.) Of or pertaining to Athens, the metropolis of Greece.⏎- (letter.) A native

  3. #3 peak

    place in a relatively small area of the Levant (which, in itself, is only a tiny part of what we today call the Middle East). Between (and including) the period of the Judges and the emergence of Israel as a kingdom,

  4. #4 peak

    6, the Coalition cabinet was authorized by the crown to present the Szell-Korber tariff to the Chamber in the form of a Hungarian autonomous tariff distinct from but identical with the Austrian tariff.⏎By an amendment of 1877,

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. L16:376 suppresses 'or' joining alternatives
  2. L13:131 excites 'and' joining parallel phrases
  3. L18:2345 suppresses
  4. L18:1047 suppresses
  5. L10:74 excites
  6. L17:447 excites
  7. L15:727 excites
  8. L9:92 suppresses "and" or "or" joining two items
  9. L12:2141 weakly excites
  10. L18:207 weakly suppresses
  11. L14:1381 weakly excites
  12. L14:2820 weakly suppresses "and" or "or" linking two counted items
  13. L15:24 weakly excites
  14. L10:1706 weakly suppresses
  15. L12:869 weakly suppresses
  16. L14:1209 weakly excites
  17. L17:35 weakly suppresses
  18. L4:57 weakly suppresses
  19. L18:509 weakly suppresses
  20. L7:122 weakly excites
  21. L4:1999 weakly suppresses
  22. L18:445 weakly suppresses
  23. L17:1044 weakly suppresses
  24. L18:693 weakly excites

Feeds later layers

  1. L22:228 strongly excites
  2. L25:9 strongly suppresses
  3. L23:30 strongly suppresses
  4. L23:127 strongly suppresses
  5. L20:586 strongly excites
  6. L20:349 strongly suppresses
  7. L27:604 strongly excites
  8. L26:318 strongly excites
  9. L23:1013 strongly excites
  10. L21:262 strongly suppresses
  11. L26:53 strongly suppresses
  12. L23:320 excites
  13. L21:930 excites
  14. L20:162 excites
  15. L20:974 excites
  16. L21:1521 excites
  17. L27:804 excites
  18. L22:518 excites
  19. L27:565 excites
  20. L23:220 excites
  21. L22:1425 suppresses
  22. L27:321 excites
  23. L22:479 excites
  24. L27:1291 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.

1 values between -14.95 and -14.051 values between -8.63 and -7.731 values between -5.93 and -5.022 values between -5.02 and -4.122 values between -4.12 and -3.225 values between -3.22 and -2.3214 values between -2.32 and -1.41155 values between -1.41 and -0.513323 values between -0.51 and 0.39258 values between 0.39 and 1.3044 values between 1.30 and 2.2023 values between 2.20 and 3.1022 values between 3.10 and 4.0016 values between 4.00 and 4.915 values between 4.91 and 5.812 values between 5.81 and 6.714 values between 6.71 and 7.625 values between 7.62 and 8.524 values between 8.52 and 9.424 values between 9.42 and 10.321 values between 10.32 and 11.232 values between 11.23 and 12.132 values between 12.13 and 13.033 values between 13.03 and 13.933 values between 13.93 and 14.841 values between 14.84 and 15.741 values between 15.74 and 16.642 values between 16.64 and 17.551 values between 17.55 and 18.451 values between 20.25 and 21.16offnegativepositiveoften

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. would
  2. could
  3. can
  4. should
  5. will
  6. may
  7. must
  8. might
  9. are
  10. is

Suppresses

  1. olate
  2. ending
  3. leaving
  4. ORS
  5. abling
  6. UGIN
  7. nable
  8. ง
  9. meaning
  10. Ỏ