Neuron Explorer

Qwen3-0.6B

the letter z inside words

A neuron in layer 27 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

The neuron fires on the letter 'z' when it appears in the middle of a word, most clearly in 'unzipping'. Its weaker opposite is the letter 'm' inside words like 'hypermiling'. Turning it up or down gave repetitive or unrelated text.

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

    ma, gzip, compress, xz, and Izop. here we will be using the .tar utility to Unzip the files.⏎The command for unzipping the .tgz file using the .tar utility:⏎tar -xv

  2. #2 peak

    Great Migration and surpassed two million occupants by 1950.⏎http://en.wikipedia.org/wiki/Philadelphia 360-degree panorama photography by yunzeng liu.⏎360-degree panorama photography by yunzeng

  3. #3 peak

    CT and Leeds University, which examined how market forces and digital technology are changing higher education, particularly in respect of emerging models of teaching and learning.⏎Professor Laura Czerniewicz, director of UCT’s Centre for Innovation in Learning and Teaching

  4. #4 peak

    a variation of zipped files; most open-source packages use this format to zip their files. If you are new to Linux or recently encountered the problem of unzipping the .tgz file, then you don’t have to worry this guide

  5. #5 peak

    with Extract Compressed Zipped Folder will appear with the zip file path.⏎Step 3: Tap on Browse and select a destination to save the files after unzipping.⏎Step 4: Enable Show Extracted Files When Complete to see the

  6. #6 peak

    de Bèze, his enthusiasm for later kabbalistic literature clearly waned. Mercier translated almost the whole of Targum Jonathan b. Uzziel on the Prophets; and he wrote annotations to Santes *P

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:uuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuu

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

    no more forgetting distant memories. Hmm …⏎Via Sustainable Business.<|endoftext|>Word of the Year: "Hypermiling"⏎For those outside the green movement, hypermiling is defined by the word-meisters of Oxford this way: "Hyperm

  2. #2 peak

    of the waning Moon IKrishna paksha) in Vedic astrology. The phase of the Moon is calculated daily in the Vedic calendar. Krishna Janmashatmi is celebrated from moonrise to midnight on the tithi of

  3. #3 peak

    School from 1883 to 1892. She excelled in her studies: in 1885 she received the Dalrymple Victoria prize; in 1888 she won an Education Board Junior

  4. #4 peak

    hypertelorism (widely-spaced eyes); (P)ulmonary stenosis (obstruction of the normal outflow of blood from the right ventricle of the heart); (A)bnormalities of the genitals; (R

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. L4:216 strongly excites
  2. L22:1766 strongly suppresses
  3. L16:0 strongly suppresses
  4. L21:196 strongly suppresses
  5. L21:690 strongly excites
  6. L19:39 strongly excites
  7. L16:10 suppresses
  8. L0:2215 suppresses
  9. L19:89 excites
  10. L2:761 suppresses
  11. L2:1083 excites
  12. L26:434 excites
  13. L2:1091 suppresses
  14. L20:2984 excites
  15. L2:801 suppresses
  16. L4:490 suppresses decade digit in years like 1930s
  17. L25:956 excites
  18. L22:271 suppresses
  19. L16:3 suppresses
  20. L16:1 suppresses
  21. L5:97 suppresses
  22. L24:980 suppresses
  23. L20:854 excites
  24. L20:961 suppresses

Feeds later layers

None: this is the last layer.

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 -11.31 and -8.811 values between -8.81 and -6.311 values between -6.31 and -3.8116 values between -3.81 and -1.313856 values between -1.31 and 1.1916 values between 1.19 and 3.694 values between 3.69 and 6.191 values between 6.19 and 8.691 values between 8.69 and 11.191 values between 11.19 and 13.692 values between 13.69 and 16.192 values between 16.19 and 18.691 values between 18.69 and 21.191 values between 21.19 and 23.691 values between 26.19 and 28.691 values between 33.69 and 36.191 values between 53.69 and 56.191 values between 86.19 and 88.69offnegativepositiveoften

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. oeff
  2. Success
  3. stile
  4. ーム
  5. Authenticated
  6. confirmed
  7. جموعة
  8. .styleable
  9. ousedown
  10. ثل

Suppresses

  1. u
  2. uC
  3. uations
  4. an
  5. ução
  6. ynomial
  7. uin
  8. m
  9. uem
  10. uí