Neuron Explorer

Qwen3-0.6B

'get', 'go' and 'got' after a subject

A neuron in layer 18 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 lights up on verbs like 'get', 'go' or 'got' that come right after a subject such as 'let's', 'it' or 'can'. When it was turned up, the model's writing drifted into repeated, rambling phrasing, so the effect on the text was not clear.

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

    .⏎There is another marker at this same location: MEEK/ELLIOTT THE TERRIBLE TRAIL<|endoftext|>Elephant Island Cruises⏎This island probably got its name after the number of Elephant seals sighted by Captain George Powell in

  2. #2 peak

    space and bandwidth while installing downloading and installing the packages.⏎This article will cover two methods of unzipping the .tgz files using the terminal. Let’s get into these.⏎Note: For this particular tutorial, we will use the Ubuntu

  3. #3 peak

    ator line benefits from its climate and agriculture production. Before you go further, it is good for you to enrich your agriculture production countries here. Now, let’s get started answering your question about why Southeast Asia is the largest agricultural production region.⏎Behind

  4. #4 peak

    Files?⏎We can zip and unzip files in Windows 10 easily. To zip files:⏎Step 1: Go to File Explorer and go to the files and media you want to zip.⏎Step 2: On an

  5. #5 peak

    Beach Park is identical in every way to the black-crowned night-heron overseas. Although this bird is native, perhaps it got here only recently, because it hasn't changed from its relatives elsewhere.⏎By contrast

  6. #6 peak

    babies come from? These questions populate our lives and inform our perceptions of the world. Explore these questions according to the expressions of artists from all over the world. Go beyond Western art to consider where art comes from and how it impacts all of us

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 part of the first part of the first part of the first... - - - - - - - - - - - - - - - -22/29.22922292222

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

    to protect yourself and your loved ones from its impact. So, let’s begin.⏎How Does Current Work?⏎Before going deeper into the concept of contact current, let’s take a moment to understand how current works.⏎Current is the rate at which

  2. #2 peak

    and fill it with water. Poke a small hole in the bottom of the bag at one of the bucket drain holes. Let the water drain out slowly near the trunk, but not on<|endoftext|>Electric cars may

  3. #3 peak

    Let’s see if you know any of these enormous structures.⏎Let’s check out these huge black holes in the video on page 2<|endoftext|>Ser

  4. #4 peak

    and teachers has been very positive. Program organizers report "many students say they're determined to visit Lake Mead NRA in the future."⏎It's a pretty<|endoftext|>Let’s go back in time, to Los Angeles in 1875.

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. L2:2029 strongly excites
  2. L16:142 strongly excites
  3. L0:2471 strongly excites
  4. L17:360 strongly excites
  5. L6:944 strongly excites
  6. L1:24 strongly suppresses
  7. L17:512 strongly excites
  8. L6:1434 strongly suppresses
  9. L7:342 strongly excites
  10. L4:711 suppresses
  11. L17:399 excites
  12. L10:1228 suppresses
  13. L16:205 excites
  14. L16:671 excites
  15. L16:374 excites
  16. L17:517 suppresses "but" in "not only X, but" phrases
  17. L16:1066 suppresses
  18. L7:219 suppresses
  19. L16:353 excites
  20. L0:690 excites
  21. L2:1585 suppresses
  22. L16:215 suppresses
  23. L15:114 excites
  24. L17:838 suppresses

Feeds later layers

  1. L20:236 strongly excites
  2. L19:690 strongly excites
  3. L25:340 suppresses
  4. L21:2864 excites
  5. L27:2364 excites
  6. L23:2836 excites
  7. L19:593 suppresses
  8. L20:256 suppresses
  9. L27:2610 excites
  10. L19:906 suppresses
  11. L19:1865 suppresses
  12. L27:2558 excites
  13. L27:2235 excites
  14. L20:2399 suppresses
  15. L27:2939 excites
  16. L27:2566 excites capital G starting a name or word
  17. L24:1837 excites
  18. L27:2695 excites
  19. L26:2719 weakly excites
  20. L19:2330 weakly suppresses
  21. L19:2417 weakly suppresses
  22. L26:1613 weakly excites
  23. L19:1004 weakly suppresses
  24. L19:594 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.

275 values between -2.14 and -1.322110 values between -1.32 and -0.511428 values between -0.51 and 0.3020 values between 0.30 and 1.118 values between 1.11 and 1.925 values between 1.92 and 2.733 values between 2.73 and 3.542 values between 3.54 and 4.362 values between 4.36 and 5.171 values between 5.17 and 5.982 values between 6.79 and 7.603 values between 7.60 and 8.412 values between 8.41 and 9.222 values between 9.22 and 10.035 values between 10.03 and 10.842 values between 10.84 and 11.662 values between 11.66 and 12.472 values between 12.47 and 13.284 values between 13.28 and 14.094 values between 14.09 and 14.901 values between 14.90 and 15.712 values between 15.71 and 16.522 values between 16.52 and 17.331 values between 17.33 and 18.155 values between 18.15 and 18.961 values between 18.96 and 19.771 values between 19.77 and 20.584 values between 20.58 and 21.393 values between 21.39 and 22.201 values between 22.20 and 23.011 values between 23.01 and 23.821 values between 23.82 and 24.631 values between 26.26 and 27.071 values between 27.07 and 27.881 values between 29.50 and 30.31offnegativepositiveoften

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. busters
  2. .TabStop
  3. bye
  4. parity
  5. olang
  6. blk
  7. illis
  8. LoggerFactory
  9. →push
  10. Braz

Suppresses

  1. usement
  2. 多く
  3. γ
  4. AndGet
  5. google
  6. uncate
  7. graph
  8. gal
  9. eu
  10. abwe