Neuron Explorer

Qwen3-0.6B

the word 'other' after 'each'

A neuron in layer 3 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 the word "other" when it comes right after "each", as in "each other" or "from each other". Turned up, it pushed the model toward counting or listing numbers, which is not obviously related.

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

    Chinese idioms, 毋忘在莒 strangely has actually two stories of origin, which gave this idiom two different meanings, though somewhat related to each other.⏎The first story of origin came from the early Spring and Autumn period. The

  2. #2 peak

    , there is a rocky giant called Mt. San Gorgonio. It is 11,502 feet tall—a half a head higher than the other peaks flanking it in the San Bernardino range. By early spring, when most

  3. #3 peak

    storm? It's good practice for talking, so encourage your baby's efforts by chatting and reading together. Your baby is also figuring out how objects relate to each other and fit together, so it's a perfect time for stacking toys. This week

  4. #4 peak

    ence electrons from the positive charge of the nucleus.⏎Electron-Electron Repulsions ⏎When two electrons are in the same shell, they will repel each other slightly. This effect is mostly canceled out due to the strong attraction to the nucleus

  5. #5 peak

    a protein to work ... . For example, if it is the job of one protein to bind specifically to a second protein, then their two shapes must fit each other like a hand in a glove. If there is a positively charged amino acid on

  6. #6 peak

    <|endoftext|>While humans prize both traits, what’s the real distinction between compassion vs empathy? Though these qualities are often used interchangeably, they’re actually different from each other. To explore the intricacies of these two emotional responses and find out how they

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: $12$ $12$ $13$ $14$ $15$ $16$ $17$ $18$ $19$ $20$ $21$ $22$

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

    04 pieces.⏎The organisation also aims to allow art enthusiasts to enjoy marvelling at the brilliant and unique Chodiev collection of Kubotas. Jacqueline Atkin, a North Carolina-based art curator and historian, is in charge of the whole

  2. #2 peak

    first idea that comes to our mind is cutting down on carbohydrate. The world is filled with various fad diets. From keto diet to paleo diet and atkin diet - almost all of these weight loss diets recommend omitting or reducing carbs from

  3. #3 peak

    , breathing exercises, pranayama (breath control), and self-control. When you have almost any queries concerning wherever along with tips on how to employ

  4. #4 peak

    34, 1650, 1733, 1703 and 1707 are included under the topic Early Cox History in all our PDF Extended History products.⏎Another 147 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. L2:36 excites
  2. L2:49 suppresses
  3. L2:111 suppresses
  4. L0:1224 suppresses
  5. L1:2801 excites
  6. L2:70 weakly excites
  7. L0:178 weakly suppresses
  8. L0:1301 weakly suppresses
  9. L2:1334 weakly suppresses
  10. L1:1878 weakly suppresses
  11. L2:1710 weakly suppresses
  12. L2:1899 weakly excites
  13. L2:2328 weakly suppresses
  14. L2:1494 weakly suppresses
  15. L2:2771 weakly excites
  16. L1:630 weakly suppresses
  17. L0:2821 weakly excites
  18. L2:75 weakly suppresses
  19. L2:366 weakly suppresses
  20. L0:987 weakly excites
  21. L2:570 weakly suppresses
  22. L2:237 weakly excites
  23. L2:126 weakly excites
  24. L0:1085 weakly excites

Feeds later layers

  1. L4:1825 suppresses
  2. L4:849 suppresses
  3. L4:2515 suppresses
  4. L6:1047 suppresses
  5. L6:1525 suppresses
  6. L9:1464 weakly suppresses
  7. L4:1302 weakly suppresses
  8. L7:1133 weakly suppresses
  9. L9:2174 weakly suppresses
  10. L6:49 weakly excites
  11. L9:42 weakly suppresses
  12. L20:870 weakly suppresses
  13. L5:3022 weakly suppresses
  14. L6:1178 weakly excites
  15. L24:1164 weakly suppresses
  16. L25:2626 weakly excites
  17. L4:2406 weakly excites
  18. L23:167 weakly suppresses
  19. L15:1056 weakly excites
  20. L10:2448 weakly excites
  21. L4:1232 weakly excites
  22. L8:24 weakly suppresses
  23. L14:2884 weakly excites
  24. L22:2164 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.

1 values between -2.73 and -2.655 values between -0.41 and -0.3321 values between -0.33 and -0.2495 values between -0.24 and -0.16390 values between -0.16 and -0.081206 values between -0.08 and 0.011281 values between 0.01 and 0.09564 values between 0.09 and 0.17230 values between 0.17 and 0.2683 values between 0.26 and 0.3422 values between 0.34 and 0.427 values between 0.42 and 0.503 values between 0.50 and 0.59offnegativepositiveoften

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. 物
  2. ь
  3. чин
  4. ее
  5. ysis
  6. chron
  7. minated
  8. 丐
  9. суд
  10. 人造

Suppresses

  1. auss
  2. itivity
  3. rowned
  4. gende
  5. grado
  6. reste
  7. avenport
  8. /full
  9. 背着
  10. قه