Neuron Explorer

Qwen3-0.6B

capital I starting a word fragment

A neuron in layer 26 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 a capital letter "I" that begins a piece of a name or word, like "Ida" or "Izyaslav", or an abbreviation like "IVM". The negative side fires on the "I" after a hyphen in names like Iran-Iraq. Steering was not clearly related: turned down the model repeated "money".

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

    is emotionless (Bhavanasunya). Manu starts getting involved in various emotions, thoughts and actions. These are sequentially portrayed with Shraddha, Ida, Kilaat and other characters playing their part, contributing in them.

  2. #2 peak

    content, not more than 5%.⏎Air-Entraining Portland Cements Specifications for four types of air-entraining portland cement (Types IA, IIA, II (MH)A, and IIIA) are given in ASTM

  3. #3 peak

    can eat several hundred mosquito larvae each day. The District uses mosquitofish as part of our Integrated Vector Management (IVM) strategy (learn more about IVM here and here). Mosquitofish are useful in non-natural sources

  4. #4 peak

    led pleasures, the impasse was inevitable. This story is so very ancient that metaphor has wonderfully mingled with history. Therefore Manu, Shraddha and Ida while maintaining their historical importance may also express the symbolic import. Manu represents

  5. #5 peak

    Kamayani depicts the interplay of human emotions, thoughts, and actions by taking mythological metaphors. Kamayani has personalities like Manu, Ida and Śraddhā who are found in the Vedas. The

  6. #6 peak

    salt consumption in humans, results in cognitive impairment. The deleterious effects of salt on the brain are independent of elevations of blood pressure," explains Dr. Iadecola.⏎According to the American Institute for Cancer Research, less than

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: money money money money money money money money money money money money money money money money money money money money money money money money money money money money money money money money money money money money money money money money money money money money money money money money

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

    atollah Khomeini attempting to act as the leader of worldwide Islam. Khomeini hoped that Shi'ites in Iraq would side with him in the Iran-Iraq War, but they did not. But many Sunnis did react against that

  2. #2 peak

    up in order to get the most out of your battery.⏎Lithium-Ion Battery: They charge faster than other types of batteries.⏎Lead Acid Battery:

  3. #3 peak

    the present.⏎By asking why the Easter Rising occurred, Ward is able to review the history of Anglo-Irish relations, from the time of Norman settlement to World War I, as well

  4. #4 peak

    Control and Prevention. Another illness linked to harmful algal blooms is<|endoftext|>The Post-Impressionist Vincent van Gogh (1853-189

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. L17:367 suppresses
  2. L17:106 excites
  3. L18:205 excites
  4. L20:121 excites
  5. L21:1492 excites
  6. L5:138 suppresses
  7. L0:47 suppresses
  8. L22:2658 excites 'I' starting an acronym or word fragment
  9. L24:586 excites
  10. L16:75 excites
  11. L16:503 suppresses
  12. L7:2095 suppresses
  13. L23:678 excites
  14. L1:45 excites
  15. L2:1923 excites
  16. L20:342 suppresses
  17. L0:2262 excites
  18. L16:84 suppresses
  19. L1:1607 excites
  20. L1:985 excites
  21. L1:14 excites
  22. L2:1463 weakly suppresses
  23. L10:1641 weakly suppresses
  24. L16:25 weakly excites

Feeds later layers

  1. L27:874 weakly suppresses
  2. L27:1226 weakly excites
  3. L27:882 weakly excites
  4. L27:458 weakly suppresses
  5. L27:1347 weakly suppresses
  6. L27:2896 weakly excites capital e starting a word
  7. L27:2099 weakly excites
  8. L27:1307 weakly excites
  9. L27:2687 weakly suppresses
  10. L27:1178 weakly suppresses
  11. L27:1981 weakly suppresses
  12. L27:2663 weakly suppresses
  13. L27:2086 weakly excites
  14. L27:185 weakly excites
  15. L27:1012 weakly suppresses
  16. L27:985 weakly excites
  17. L27:1834 weakly suppresses
  18. L27:1724 weakly suppresses
  19. L27:1042 weakly excites
  20. L27:2485 weakly excites
  21. L27:961 weakly suppresses
  22. L27:1982 weakly excites
  23. L27:2491 weakly suppresses
  24. L27:1531 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 -22.09 and -19.981 values between -15.76 and -13.652 values between -11.54 and -9.423 values between -9.42 and -7.316 values between -7.31 and -5.2010 values between -5.20 and -3.0950 values between -3.09 and -0.983771 values between -0.98 and 1.1447 values between 1.14 and 3.257 values between 3.25 and 5.364 values between 5.36 and 7.471 values between 7.47 and 9.581 values between 11.69 and 13.811 values between 15.92 and 18.031 values between 20.14 and 22.251 values between 22.25 and 24.361 values between 60.26 and 62.38offnegativepositiveoften

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. weise
  2. tableFuture
  3. cular
  4. tlement
  5. iola
  6. Await
  7. bery
  8. isecond
  9. 我都
  10. rients

Suppresses

  1. g
  2. シン
  3. tert
  4. nio
  5. owa
  6. 从此
  7. nie
  8. 마련
  9. otos
  10. gne