Neuron Explorer

Qwen3-0.6B

"one of" before a group

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 fires on "of" in phrases like "one of sixty images" or "one of the fastest", where something is being singled out from a larger group. When turned up, the model kept writing "many" and "a few".

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

    Voyager mission⏎The pale blue dot image was taken on 14 February 1990. It was just one of sixty images taken that day that were designed to capture most of the planets of the

  2. #2 peak

    solar cells that soldiers can roll up like a mat, transport, and unroll in a new location to start generating electricity on the spot.⏎It’s only one of the technologies South Dakota State University is exploring as part of a $10 million

  3. #3 peak

    And yet, experts predict more than 1 million computing jobs will be created in the next 10 years, according to the website.⏎Deister is one of the teachers, professionals, students and administrators nationwide pushing to increase access to computer programming

  4. #4 peak

    ite hit atmosphere at 64,000 mph⏎A meteorite that exploded as a fireball over California's Sierra foothills last April was one of the fastest and rarest known to have hit Earth.⏎A minivan-sized carbon

  5. #5 peak

    a number of cancers and scientists around the globe are busy trying to harness the power of the immune system in the fight against cancer. The HPV vaccine is just one of the first; expect to hear more and more about the topic in the future.⏎

  6. #6 peak

    The pair laid three eggs this year in February. None of their eggs hatched last year, the DNR said. The "EagleCam" is one of the program's projects funded by Minnesot

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 first element of the first element, the first element of the first element, the first element of the first first element, the first first first, the first first first first, the first first first first first, etc. What is the

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

    likely to develop psychiatric symptoms later on, and vice versa. Treatment for sleep disorders in young children may limit the development of psychiatric disorders later on.⏎The researchers interviewed the parents of nearly 1,000 toddlers. The toddlers were part of

  2. #2 peak

    wondered if he really died because I never got to see and hold his body."⏎The disease hit the DRC in August 2018 and has claimed the lives of over 2,200 people, according to the World Health

  3. #3 peak

    bursts and cosmic explosions, but it turns out that the spacecraft has been involved in 10 asteroid-observing sessions as well.⏎Swift's scientists had the satellite watch a couple of patches of sky that YU55 was predicted to

  4. #4 peak

    00 non-words.⏎To check whether the pigeons were learning to distinguish words from non-words rather than merely memorising them, the researchers introduced words the birds had never seen before.⏎The pigeons correctly identified the new words as words

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. L17:30 suppresses
  2. L10:1377 suppresses
  3. L16:792 excites "the" inside a prepositional phrase
  4. L9:24 excites "the" before a named organisation or title
  5. L11:1168 excites
  6. L9:662 suppresses
  7. L17:831 suppresses
  8. L4:52 excites
  9. L10:120 suppresses
  10. L16:419 excites
  11. L13:1234 suppresses
  12. L11:875 suppresses
  13. L4:73 suppresses
  14. L16:2248 suppresses
  15. L18:1547 suppresses
  16. L18:171 excites
  17. L18:602 suppresses
  18. L15:1611 weakly suppresses
  19. L12:435 weakly suppresses
  20. L6:61 weakly suppresses
  21. L10:1162 weakly excites
  22. L4:217 weakly suppresses
  23. L5:2953 weakly suppresses
  24. L15:1848 weakly suppresses

Feeds later layers

  1. L21:822 strongly excites
  2. L22:825 strongly excites 'for' before a span of time
  3. L27:810 strongly excites
  4. L23:538 excites
  5. L21:2090 excites
  6. L20:2136 excites
  7. L23:2230 excites
  8. L24:2713 excites 'a' before few or variety
  9. L20:2987 excites
  10. L20:1935 excites
  11. L27:932 excites
  12. L26:1784 excites
  13. L20:2316 excites
  14. L24:1848 excites
  15. L21:785 excites
  16. L23:96 suppresses
  17. L20:2745 excites
  18. L23:870 excites
  19. L22:1358 excites
  20. L24:826 excites
  21. L20:1788 excites
  22. L27:545 excites
  23. L22:926 weakly excites
  24. L26:1083 weakly 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 -16.97 and -15.951 values between -9.82 and -8.806 values between -8.80 and -7.785 values between -7.78 and -6.769 values between -6.76 and -5.7415 values between -5.74 and -4.7230 values between -4.72 and -3.7025 values between -3.70 and -2.6843 values between -2.68 and -1.6671 values between -1.66 and -0.643241 values between -0.64 and 0.38389 values between 0.38 and 1.4028 values between 1.40 and 2.4313 values between 2.43 and 3.457 values between 3.45 and 4.4710 values between 4.47 and 5.493 values between 5.49 and 6.512 values between 6.51 and 7.531 values between 7.53 and 8.551 values between 8.55 and 9.571 values between 9.57 and 10.592 values between 11.61 and 12.631 values between 12.63 and 13.651 values between 13.65 and 14.671 values between 14.67 and 15.691 values between 22.84 and 23.86offnegativepositiveoften

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. many
  2. several
  3. few
  4. 许多
  5. 很多
  6. 多种
  7. 若干
  8. Several
  9. 几种
  10. many

Suppresses

  1. himself
  2. "#{
  3. 也不例外
  4. 他也
  5. الآخر
  6. ("#{
  7. NDEBUG
  8. 做到
  9. lick
  10. 否则