Neuron Explorer

Qwen3-0.6B

unclear: time words like 'later' and 'years'

A neuron in layer 15 of Qwen3-0.6B. Its strongest snippets are mixed: they peak on different tokens. It is featured, so it has steering outputs.

What this neuron does

This neuron fires on a mix of words that often end a short phrase, such as 'later', 'years', 'day' and 'page'. It is hard to find one clear theme, and turning it up gave rambling, off-topic text.

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

    of liquid as a bead, like a bead of sweat on your forehead. The earliest meaning of bead is "prayer bead," from the Old English gebed, "prayer."<|endoftext|>Solar geoengineering (SG) and CO2 emissions reduction

  2. #2 peak

    hard drive in the same PC.⏎Note: always be cautious when using dd, as it’s very easy to completely wipe out a drive, as we will show later in this article.⏎Creating a Drive Image⏎Boot up into the Ubuntu Live CD

  3. #3 peak

    directly behind his backyard in 1889.⏎Bailey can be seen sitting in that forcing house in the cyanotype photograph reproduced at the top of this page<|endoftext|>August 21, 2015⏎Station name board in

  4. #4 peak

    (92 mother tongues) The majority of the population are Hindu Adult Literacy Rate (6 and over): 53% Life expectancy: 68 years Home of Mt. Everest and the highest altitude lake in the world (Til

  5. #5 peak

    for Nurses in New York.⏎1916-1918: Mary Breckinridge’s newborn daughter, Polly, dies, followed two years later by the unexpected death of her four-year-old son, Breckie. The

  6. #6 peak

    blind, or homebound person in a hospital or in an extended-care⏎- Perform volunteer<|endoftext|>Nanocrystals May Provide Boost for Solar Cells, Solar Hydrogen Production⏎5 January 2006⏎Los Alamos National Laboratory scientists

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: I think I was thinking about the same thing, but I was thinking of the same thing. So, I think that was something, but I was thinking of the same thing. I think I was thinking of the same thing. So,

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

    2 Diabetes⏎Wheat bran cereal is one of the richest plant sources of magnesium. This mineral supports the activities of more than 300 enzymes, including enzymes involved in secretion of insulin and utilization of glucose. Studies have established that wheat bran

  2. #2 peak

    the U.S. Constitution.⏎In this year’s Religious Freedom Day proclamation, President Barack Obama writes,⏎Today, America embraces people of all faiths and of no faith. We are Christians and Jews, Muslims and Hindus, Buddhists and Sikhs

  3. #3 peak

    , which specifies where the runtime can locate an assembly (for more information, see Specifying an Assembly's Location).⏎<codeBase version="188.8.131.52" href="http://www.litware

  4. #4 peak

    children are naturally independent, with a strong need to express their own will. They test any boundaries set for them, including those set by their parents. These children can grow up to be rule-breakers, creative

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. L3:294 suppresses
  2. L8:10 suppresses
  3. L8:7 suppresses
  4. L13:23 weakly suppresses
  5. L14:57 weakly suppresses
  6. L9:70 weakly excites
  7. L14:0 weakly suppresses
  8. L14:252 weakly excites
  9. L3:59 weakly suppresses
  10. L6:8 weakly excites
  11. L12:625 weakly excites
  12. L12:110 weakly suppresses
  13. L11:1536 weakly excites
  14. L12:940 weakly excites
  15. L12:1835 weakly suppresses
  16. L1:87 weakly excites
  17. L1:2990 weakly excites
  18. L7:62 weakly suppresses
  19. L12:2289 weakly excites 'of' after part-whole nouns
  20. L7:102 weakly excites
  21. L13:122 weakly suppresses
  22. L14:597 weakly suppresses unclear: document ends and list breaks
  23. L5:169 weakly suppresses
  24. L14:440 weakly suppresses

Feeds later layers

  1. L18:319 suppresses
  2. L16:309 weakly suppresses
  3. L19:2883 weakly excites
  4. L17:1985 weakly excites
  5. L18:2681 weakly excites
  6. L22:299 weakly excites
  7. L17:203 weakly suppresses
  8. L19:2542 weakly excites
  9. L20:2813 weakly excites
  10. L16:775 weakly suppresses
  11. L16:2599 weakly suppresses
  12. L16:412 weakly suppresses
  13. L18:1613 weakly excites
  14. L20:1979 weakly suppresses
  15. L23:1931 weakly excites
  16. L26:1803 weakly suppresses
  17. L21:1888 weakly excites
  18. L16:2074 weakly excites
  19. L25:184 weakly excites
  20. L16:1049 weakly suppresses
  21. L21:2678 weakly excites
  22. L16:1011 weakly suppresses
  23. L16:2486 weakly suppresses
  24. L18:549 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 -3.98 and -3.792 values between -3.60 and -3.413 values between -3.03 and -2.843 values between -2.84 and -2.651 values between -2.65 and -2.456 values between -2.45 and -2.2611 values between -2.26 and -2.0714 values between -2.07 and -1.8815 values between -1.88 and -1.6926 values between -1.69 and -1.5043 values between -1.50 and -1.3161 values between -1.31 and -1.1283 values between -1.12 and -0.93102 values between -0.93 and -0.74167 values between -0.74 and -0.55246 values between -0.55 and -0.36313 values between -0.36 and -0.17449 values between -0.17 and 0.02496 values between 0.02 and 0.21452 values between 0.21 and 0.40374 values between 0.40 and 0.59295 values between 0.59 and 0.79229 values between 0.79 and 0.98171 values between 0.98 and 1.17113 values between 1.17 and 1.3684 values between 1.36 and 1.5555 values between 1.55 and 1.7442 values between 1.74 and 1.9319 values between 1.93 and 2.129 values between 2.12 and 2.3113 values between 2.31 and 2.502 values between 2.50 and 2.694 values between 2.69 and 2.883 values between 2.88 and 3.071 values between 3.45 and 3.64offnegativepositiveoften

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. /light
  2. .neo
  3. /Internal
  4. НЕ
  5. ynet
  6. 说了
  7. ologue
  8. ""; ⏎
  9. kest
  10. egin

Suppresses

  1. orque
  2. urally
  3. isque
  4. oque
  5. incare
  6. arkadaş
  7. 队员们
  8. koş
  9. Separator
  10. ッ