Neuron Explorer

Qwen3-0.6B

word fragment starting a longer name

A neuron in layer 24 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 fires on the first syllables of a longer word, especially names like 'Zo' in Zoanthidae or 'Og' in Ogden. Turning it up made the model repeat the word 'over' again and again.

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

    colonial anthozoans and soft coral by their characteristic of incorporating sand and other small pieces of material into their tissue to help make their structure (except for the family Zoanthidae). A main characteristic of the order is that their tentacles are

  2. #2 peak

    ite blog about Israeli museums’ spotlight on Circassians.⏎“We’ve traditionally been friends with both Jews and Arabs, even when they were fighting among themselves,” said Zoher Thawcho, the 39-year-old co-founder of the Circ

  3. #3 peak

    by total area.⏎The county extends from high in the Wasatch Range in the east into a portion of the Great Salt Lake to the west. The Weber and Ogden rivers and their tributaries run through its valleys. The Weber County Survey

  4. #4 peak

    this was an easy matter and a parergon. They must apply themselves zealously and in the deepest secrecy to the study

  5. #5 peak

    anthids are the zooxanthellate genera found in tropical and sub-tropical waters (primarily Zoanthus and Palythoa), many other species and genera exist, some

  6. #6 peak

    B" and "C" Batteries⏎saw plenty of action and subdued several machine gun nests that were holding⏎up the infantry, whose initial pace from Egret trench

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 main source of each person from the r class has to be from the r class each person each r class each person each person from the r class each r class each person each r class each person each r class each person each r class

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

    data. Further, the cloud and artificial intelligence have a symbiotic relationship. The cloud handles the fundamental source of information that augments artificial intelligence networks, and artificial<|endoftext|>If I were to think of an emblem

  2. #2 peak

    finding, the researchers have clarified the way in which this oscillator adjusts its activity throughout the day. They show that the three proteins studied, the Pseudo-Response Regulators PRR5, PRR7 and PRR9, associate with promoter

  3. #3 peak

    . To the region’s most ardent segregationists, the Douglas DC-4s and Lockheed Electras landing from faraway cities on shimmering southern runways augured a<|endoftext|>Canada's tar sands, located in the province of Alberta, are

  4. #4 peak

    Bhushan, who inaugurated the conference, said the philosophy of sustainable development required maintaining a balance between the impact of anthropogenic activities and the ecosystem.⏎Inaugurating the conference, Anna University Vice-Chancellor

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. L23:2981 suppresses
  2. L2:1632 weakly suppresses
  3. L22:384 weakly excites
  4. L6:2398 weakly excites
  5. L23:1489 weakly excites
  6. L19:1004 weakly suppresses
  7. L22:715 weakly suppresses
  8. L15:2362 weakly excites
  9. L2:1774 weakly suppresses
  10. L8:1506 weakly suppresses
  11. L17:360 weakly excites
  12. L9:923 weakly excites
  13. L8:418 weakly excites
  14. L11:1031 weakly excites
  15. L2:1229 weakly excites
  16. L19:2271 weakly excites
  17. L18:1350 weakly suppresses
  18. L9:213 weakly excites
  19. L16:1236 weakly suppresses
  20. L19:1052 weakly excites
  21. L2:2255 weakly suppresses
  22. L13:1820 weakly excites
  23. L18:2383 weakly excites
  24. L15:2116 weakly excites

Feeds later layers

  1. L26:911 weakly excites
  2. L25:1686 weakly suppresses
  3. L26:1505 weakly excites
  4. L27:75 weakly suppresses
  5. L25:126 weakly excites
  6. L26:2131 weakly suppresses
  7. L27:2251 weakly suppresses
  8. L27:285 weakly suppresses
  9. L26:2949 weakly excites
  10. L27:1977 weakly suppresses
  11. L27:1044 weakly excites
  12. L25:332 weakly excites
  13. L25:2625 weakly suppresses
  14. L25:619 weakly suppresses unclear: word fragments and 'or'/'and' in lists
  15. L27:3041 weakly suppresses
  16. L25:1256 weakly excites power and energy unit symbols
  17. L25:1468 weakly excites
  18. L27:945 weakly suppresses
  19. L27:1780 weakly suppresses
  20. L27:1294 weakly excites
  21. L25:402 weakly excites
  22. L27:455 weakly excites
  23. L26:2552 weakly excites
  24. L25:1788 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 -18.45 and -17.121 values between -17.12 and -15.793 values between -14.45 and -13.123 values between -11.79 and -10.454 values between -10.45 and -9.123 values between -9.12 and -7.793 values between -7.79 and -6.451 values between -6.45 and -5.1212 values between -5.12 and -3.7916 values between -3.79 and -2.4644 values between -2.46 and -1.122874 values between -1.12 and 0.21818 values between 0.21 and 1.5561 values between 1.55 and 2.8820 values between 2.88 and 4.2116 values between 4.21 and 5.558 values between 5.55 and 6.883 values between 6.88 and 8.212 values between 8.21 and 9.546 values between 9.54 and 10.881 values between 10.88 and 12.212 values between 12.21 and 13.541 values between 13.54 and 14.882 values between 16.21 and 17.542 values between 17.54 and 18.881 values between 33.54 and 34.88offnegativepositiveoften

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. 过
  2. -over
  3. over
  4. over
  5. 過
  6. Over
  7. Over
  8. .over
  9. _over
  10. ancement

Suppresses

  1. ASF
  2. utzer
  3. [pos
  4. lus
  5. hazırl
  6. imshow
  7. 非要
  8. asper
  9. Manor
  10. ivr