Neuron Explorer

Qwen3-0.6B

'and' joining two actions

A neuron in layer 17 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 'and' when it links two actions in a row, like 'packed up our gear and selected food'. Turning it up made the model produce repeated filler words, so the steering does not clearly show this role.

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

    Albemarle County was subsequently enlarged, and in 1696 the area south of Albemarle Sound was removed from Albemarle and made into a new county, named Bath

  2. #2 peak

    own language. By the 8th century, the Cham script had outgrown Sanskrit and the Cham language was in full use. Most preserved manuscripts focus on religious rituals,

  3. #3 peak

    potential gentrification that often comes with neighborhood investment in disadvantaged communities.⏎The Urban Green<|endoftext|>Every winter, California newts leave the safety of their forest burrows and travel as far as three miles to mate in the pond where they were born.

  4. #4 peak

    surface-to-air missile.⏎Powers released his parachute and floated down to the surface where he was detained by KGB agents and interrogated

  5. #5 peak

    thousands of Japanese American internees. One of the first families to arrive was the Wakatsukis, who were ordered to leave their fishing business in Long Beach and take with them only the belongings they could carry. For Jeanne Wakatsuki

  6. #6 peak

    are off limits (unless a dieter chooses to put the risk of severe foodborne illness aside and eat raw meat). For this reason, the raw food diet typically resembles a vegan

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: What if the one thing that is, in the one, that one, that one, that one, that one, that one, that one, that one, that one, that one, that one, that one, that one,

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

    up to 100 times faster than Map Reduce for iterative algorithms or interactive data mining.⏎- Learn how it provides in-memory cluster computing for lightning fast speed and supports Java, Python, R, and Scala APIs for ease of development.⏎-

  2. #2 peak

    Germany occurred in 1966, but it was the oil crisis of 1973/74 that ended the "economic miracle" once and for all. By the time Herbert Grünewald succeeded Kurt Hansen as Management Board

  3. #3 peak

    pupils are taught and governed, are what it mainly embraces, and the teacher is successful in the Art of Teaching, who trains children to be symmetrically developed men and women, in a physical, intellectual and moral sense. Certain modes of procedure,

  4. #4 peak

    transported soil from the mainland, tucking it into crevices on this otherwise barren rock in hopes of cultivating flowers and vegetables. For a few months in summer and early fall, the rock became a colorful garden in the

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. L10:155 excites
  2. L15:184 excites
  3. L13:44 suppresses inside web addresses and citations
  4. L13:131 excites 'and' joining parallel phrases
  5. L16:376 suppresses 'or' joining alternatives
  6. L11:223 excites
  7. L8:328 excites
  8. L4:76 excites
  9. L15:59 suppresses
  10. L16:1184 excites
  11. L7:92 excites
  12. L14:1639 suppresses "and" joining two nouns
  13. L10:74 excites
  14. L2:24 suppresses
  15. L5:548 excites
  16. L12:158 suppresses
  17. L16:771 suppresses
  18. L9:92 weakly suppresses "and" or "or" joining two items
  19. L15:830 weakly excites
  20. L4:57 weakly suppresses
  21. L10:370 weakly suppresses
  22. L15:1260 weakly excites
  23. L10:1377 weakly suppresses
  24. L15:2765 weakly excites

Feeds later layers

  1. L19:1769 excites
  2. L21:299 excites
  3. L19:271 excites
  4. L20:1214 excites
  5. L19:1608 excites
  6. L25:824 excites
  7. L18:2045 excites
  8. L21:535 weakly excites
  9. L18:472 weakly excites
  10. L18:715 weakly excites
  11. L24:398 weakly excites
  12. L23:566 weakly excites
  13. L18:529 weakly excites
  14. L20:1124 weakly suppresses
  15. L18:1533 weakly excites
  16. L19:385 weakly suppresses
  17. L19:652 weakly excites
  18. L18:2393 weakly excites
  19. L20:2815 weakly excites
  20. L20:1399 weakly excites
  21. L18:2216 weakly excites
  22. L18:740 weakly excites
  23. L19:504 weakly suppresses
  24. L18:597 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.

2 values between -2.06 and -1.842 values between -1.84 and -1.611 values between -1.61 and -1.381 values between -1.38 and -1.153 values between -1.15 and -0.922 values between -0.92 and -0.7027 values between -0.70 and -0.47200 values between -0.47 and -0.241095 values between -0.24 and -0.011809 values between -0.01 and 0.22638 values between 0.22 and 0.4486 values between 0.44 and 0.678 values between 0.67 and 0.904 values between 0.90 and 1.136 values between 1.13 and 1.352 values between 1.35 and 1.583 values between 1.58 and 1.811 values between 1.81 and 2.041 values between 2.04 and 2.261 values between 2.26 and 2.491 values between 2.49 and 2.724 values between 2.72 and 2.951 values between 3.40 and 3.631 values between 3.86 and 4.092 values between 4.09 and 4.312 values between 4.31 and 4.541 values between 4.77 and 5.002 values between 5.00 and 5.231 values between 5.68 and 5.911 values between 6.82 and 7.05offnegativepositiveoften

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. 堃
  2. ędzi
  3. øy
  4. azio
  5. readcrumbs
  6. dependencies
  7. VERAGE
  8. charts
  9. schläge
  10. 朴

Suppresses

  1. them
  2. issent
  3. They
  4. loro
  5. ihnen
  6. .Next
  7. →dc
  8. Next
  9. 家都知道
  10. "].(