Neuron Explorer

Qwen3-0.6B

"with" after a group noun

A neuron in layer 18 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

The neuron fires on the word "with" when it follows a group of people or things, as in "countries with" or "people with". It goes strongly negative on "with" in phrases like "with confidence" or "with permission". Turning it up gave mixed, unrelated text.

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

    . From Flint Michigan and California in the United States of America to Chennai, India and Cape Town, South Africa, 12 out of 17 countries with the highest water stress are in the Middle East and North Africa.⏎Flooding,

  2. #2 peak

    2002 and 11.7 percent since 1997, the analysis shows.⏎Texas tops the list of the 10 states with the biggest one-year increases in CO2 emissions, with Georgia, Arizona, California

  3. #3 peak

    need to plant 6,631,898 trees in order to offset trips for less than 1 mile per year.⏎(Of course for anyone with mobility issues, these car journeys are necessary; therefore, the analysis accounted for the

  4. #4 peak

    everything run more efficiently is an amazing adaptation. And for many people in the modern world this is still a lifesaving mechanism. But for the people in the world with easy access to Coca-Cola and McDonald’s its making it harder to stay at a

  5. #5 peak

    slightly higher in men, because they tend to spend more time in the sun and use less sun protection than women do. African-Americans, Hispanics, Asians and others with darker skin are not as susceptible as Caucasians (Skin Cancer and Skin of Color

  6. #6 peak

    for the following reasons:⏎- African Americans still reside mainly in separate and unequal communities. In 2010, in the 100 metropolitan areas with the largest African American populations, 62.5 percent of blacks would have

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 purpose of a bit of care, as in the care of as much frequency precision as possible, in the purpose of precision precision, as in the precision of the frequency as precision, as in the precision of a precision, as in the

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

    ’s book appeared, many other anthropologists have declared their acceptance of the theory, so that to-day we may assert with confidence that the black man is black because of the excessive sunlight of his environment,

  2. #2 peak

    , without safety screening⏎Reproduced with permission from Fluechtling.net⏎- Millions of refugees guaranteed German standard of

  3. #3 peak

    the biggest enemies of our health unless they are eaten very sparingly.⏎Nutrition-health relationships have been established for decades as highly related so that today we can say with certainty that eating habits and life are critical to promoting health and by contrast, are

  4. #4 peak

    herdsmen began storing milk and discovered (probably by accident) that fermented milk was safe to consume.⏎With such an extensive backstory, it’s impossible to know with certainty who invented yogurt. One thing is for sure: the history of yogurt and

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. L17:2363 suppresses
  2. L15:1162 suppresses
  3. L2:863 weakly suppresses
  4. L7:1056 weakly excites
  5. L9:31 weakly suppresses
  6. L4:882 weakly suppresses 'with' after provide or familiarize
  7. L2:2889 weakly excites
  8. L4:228 weakly suppresses
  9. L16:1860 weakly suppresses
  10. L9:1856 weakly excites "with" after a noun, introducing contents
  11. L11:1108 weakly suppresses
  12. L12:73 weakly suppresses
  13. L0:2342 weakly suppresses
  14. L15:986 weakly excites
  15. L6:738 weakly suppresses
  16. L15:1913 weakly suppresses
  17. L16:364 weakly excites
  18. L13:1087 weakly suppresses
  19. L6:2708 weakly suppresses
  20. L2:45 weakly excites
  21. L14:2116 weakly suppresses
  22. L17:964 weakly suppresses
  23. L1:1188 weakly excites
  24. L2:1321 weakly suppresses

Feeds later layers

  1. L22:2587 suppresses
  2. L23:2881 suppresses
  3. L19:1869 suppresses
  4. L19:1818 suppresses
  5. L26:1318 suppresses
  6. L19:1739 suppresses
  7. L19:2407 suppresses
  8. L20:1733 suppresses
  9. L20:1501 suppresses
  10. L22:1823 weakly suppresses
  11. L25:2637 weakly excites
  12. L22:809 weakly suppresses
  13. L24:521 weakly suppresses
  14. L23:142 weakly excites
  15. L24:921 weakly excites
  16. L19:879 weakly excites
  17. L21:2352 weakly suppresses
  18. L20:1900 weakly excites
  19. L19:1871 weakly excites
  20. L23:1587 weakly suppresses
  21. L25:799 weakly suppresses
  22. L19:2723 weakly suppresses 'of' after a quality or importance adjective
  23. L22:507 weakly suppresses
  24. L22:2814 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 -21.14 and -20.461 values between -13.01 and -12.331 values between -10.98 and -10.301 values between -7.59 and -6.911 values between -5.56 and -4.881 values between -4.88 and -4.202 values between -3.52 and -2.842 values between -2.84 and -2.174 values between -2.17 and -1.495 values between -1.49 and -0.81143 values between -0.81 and -0.133388 values between -0.13 and 0.54338 values between 0.54 and 1.2218 values between 1.22 and 1.901 values between 1.90 and 2.581 values between 5.29 and 5.97offnegativepositiveoften

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. 一族
  2. спект
  3. 诱人
  4. !$
  5. прод
  6. 三个方面
  7. 单身
  8. cased
  9. 其次
  10. –⏎⏎

Suppresses

  1. precision
  2. ease
  3. certainty
  4. precision
  5. 精度
  6. difficulty
  7. efficiency
  8. accuracy
  9. outh
  10. Precision