Neuron Explorer

Qwen3-0.6B

the 'un' prefix in words

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 the prefix 'un' in the middle of a word, such as unless, ungodly, unselfish and unimportant. Turning it up made the model produce repeated, garbled text, so the steering does not show a clear meaning.

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

    (Mt. 12: 41).<|endoftext|>“There shall never be any bond slaverie, villinage or Captivitie amongst us unles it be lawfull Captives taken in just warres, and such strangers

  2. #2 peak

    will rule in the fear of God.’⏎The preservation of a republican government depends on the faithful discharge of this duty.⏎If the citizens neglect their duty and place unprincipled men in office, the government will soon be corrupted; laws will

  3. #3 peak

    of the second death in the "lake of fire" which will befall the ungodly.⏎Gehenna, or the Valley of Hinnom, was

  4. #4 peak

    bad with most characteristics exaggerated.The hero and heroine are usually young.The heroine is usually fair, kind, charitable, and caring.The hero is usually honorable, courageous, unselfish, and caring.Both usually have special abilities or powers.⏎What is

  5. #5 peak

    entices us to go deeper and reflect upon the meaning of things, to imagine how our ancestors conceived human nature and to think about what they thought was dangerous and unfruitful and why they felt the need to issue an advice and a warning for

  6. #6 peak

    of the skiers. It has been a peculiar battle. Both sides feel strongly on the matter, but even in the most crucial moments, there has been little uncontained anger. Indeed, the only thing that has been expended at all recklessly

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: Imblebleblebleblebleblebleblebleblebleblebleblebleblebleblebleblebleblebleblebleblebleblebleblebleblebleblebleblebleblebleblebleblebleblebleble

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

    the time of electrophysiologic testing . Type I atrial flutter was classified as a macroreentrant atrial tachycardia (see "Electrocardiographic and electrophysiologic features of atrial flutter"), while type

  2. #2 peak

    also include tests for infections (microbiology), changes in body tissue and organs (anatomical histology), and chemicals, eg medication, drugs or poisons (toxicology and pharmacology).⏎In Adelaide, the post-mortem is performed

  3. #3 peak

    of the cause can be much more difficult.⏎There are some common triggers for hives. While it varies from individuals, these may include:⏎- Allergens (e.g., foods, pollens, animals, insect stings).⏎- Irr

  4. #4 peak

    As a result, scientists began to devote time to searching for drugs that would kill these disease-causing bacteria. The goal of such research was to find so-called “magic bullets” that would destroy microbes without toxicity to the person taking the drug.⏎

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. L24:483 excites
  2. L23:864 suppresses
  3. L25:239 excites
  4. L20:605 suppresses
  5. L20:563 excites
  6. L24:620 excites
  7. L23:80 excites blank slot before a year or number
  8. L23:225 suppresses
  9. L21:1291 suppresses
  10. L21:1014 excites
  11. L21:959 suppresses
  12. L23:1015 suppresses
  13. L21:390 suppresses
  14. L15:59 weakly excites
  15. L19:305 weakly suppresses
  16. L21:118 weakly suppresses
  17. L20:267 weakly suppresses
  18. L8:328 weakly suppresses
  19. L22:245 weakly suppresses
  20. L20:270 weakly suppresses
  21. L23:196 weakly excites em dash opening an aside
  22. L17:1821 weakly excites
  23. L22:54 weakly suppresses
  24. L16:588 weakly suppresses

Feeds later layers

  1. L27:1385 weakly excites
  2. L27:1448 weakly suppresses
  3. L27:2774 weakly excites
  4. L27:2073 weakly excites
  5. L27:941 weakly suppresses
  6. L27:1273 weakly suppresses
  7. L27:1160 weakly suppresses word starts of acronyms and names
  8. L27:2941 weakly excites
  9. L27:1247 weakly suppresses
  10. L27:1399 weakly suppresses
  11. L27:2720 weakly excites
  12. L27:3051 weakly excites
  13. L27:1777 weakly excites
  14. L27:302 weakly suppresses
  15. L27:2160 weakly excites
  16. L27:320 weakly excites
  17. L27:402 weakly suppresses
  18. L27:538 weakly excites
  19. L27:1658 weakly excites
  20. L27:928 weakly excites
  21. L27:2152 weakly excites
  22. L27:1066 weakly suppresses
  23. L27:2474 weakly suppresses
  24. L27:1403 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.

2 values between -28.45 and -26.951 values between -22.43 and -20.921 values between -20.92 and -19.421 values between -19.42 and -17.912 values between -17.91 and -16.414 values between -16.41 and -14.902 values between -14.90 and -13.406 values between -13.40 and -11.898 values between -11.89 and -10.386 values between -10.38 and -8.889 values between -8.88 and -7.3727 values between -7.37 and -5.8729 values between -5.87 and -4.3651 values between -4.36 and -2.85116 values between -2.85 and -1.352354 values between -1.35 and 0.161194 values between 0.16 and 1.6659 values between 1.66 and 3.1713 values between 3.17 and 4.6812 values between 4.68 and 6.182 values between 6.18 and 7.694 values between 7.69 and 9.191 values between 10.70 and 12.212 values between 12.21 and 13.711 values between 13.71 and 15.221 values between 30.28 and 31.78offnegativepositiveoften

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. ][
  2. mith
  3. rength
  4. amental
  5. ocrin
  6. tes
  7. lover
  8. loy
  9. matic
  10. roc

Suppresses

  1. ble
  2. LY
  3. Im
  4. 让孩子
  5. BODY
  6. im
  7. arin
  8. argin
  9. 把这些
  10. _cast