Neuron Explorer

Qwen3-0.6B

the word 'more' before a comparison

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

This neuron responds to the word 'more', especially in phrases like 'more than', 'more serious' or 'more recent'. The negative side is fragments of place names and words split in the middle. Steering the neuron did not produce a clear change in the 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

    plastic ― into one general area. However, it does not converge into a floating island of trash, like some people have come to believe.⏎Instead, it’s more like a cloud of pollution, and much of it is made up of particles called

  2. #2 peak

    volume laying the groundwork for his thesis that mathematics relies on a foundation constructed on assumptions. Along the way, he references certain principles of statistics and calculus as well as more esoteric branches of mathematics. He also touches on the work of several famous mathematic

  3. #3 peak

    its imaging infrastructure was seen as a benchmark for European-wide cooperation in imaging, leading to the decision to place the administrative centre in Turku.⏎The consortium builds on more than a decade of cooperation between dozens of

  4. #4 peak

    nose are fragile, so it is no wonder that they can easily bleed. However, sometimes nose bleeds can become frequent. This may be an indication of a more serious problem.⏎One of the most common causes of frequent nose bleeds is dry

  5. #5 peak

    learning anywhere, anytime; and responsive to the 21st century context of an interconnected global economy and rapidly emerging global society?⏎That was an organizing question for more than a thousand people who gathered in Doha, Qatar, for the inaugural World

  6. #6 peak

    a Minute ” was interesting and had this fascinating graphic comparing price changes from Sears, Amazon and Best Buy. In addition, this academic article explains yield management while this more recent study looks at internet dynamic pricing.<|endoftext|>A diplomatic crisis is engulfing part

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: "This is a story... it's a story... this is... this... this... this... this... this..." but... this... this... this... this... this..." and... this... this... this..." from and and and

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

    like church and tax books being destroyed by the flood, like the land taxes in the year 1731 from the villages Meedhuizen and Slochteren and the church books before 1700.⏎Graphics by Lynn Water

  2. #2 peak

    Luther's preaching and writing sparked the Protestant Reformation, which spread through much of Europe, including France. Inspiration pour des voyages à Londres, en Angleterre, en Ecosse, au Pays de Galles et en Irland

  3. #3 peak

    of York, artisans, £10 million grant, constructing a chapel, craft apprenticeships, Danish invaders, Decorated Gothic, Decorated Gothic nave, doleritic sedimentary rock, Early English, Early English north and south transepts

  4. #4 peak

    , of which only three have survived: the⏎Auberge de France, the Auberge d'Angleterre and the Auberge de Provence and⏎After the foundation of Vallet

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. L5:858 strongly excites
  2. L1:47 strongly excites
  3. L16:139 strongly suppresses
  4. L0:41 suppresses
  5. L16:463 suppresses 'for' introducing a purpose or target
  6. L17:253 suppresses
  7. L16:66 suppresses
  8. L16:113 excites
  9. L0:2211 suppresses
  10. L2:1866 suppresses
  11. L15:96 suppresses
  12. L3:1055 suppresses
  13. L16:60 suppresses
  14. L1:32 excites
  15. L0:2765 excites
  16. L14:95 excites
  17. L0:2724 suppresses
  18. L10:1699 weakly excites
  19. L17:1637 weakly excites
  20. L4:189 weakly excites
  21. L17:725 weakly suppresses
  22. L4:344 weakly suppresses
  23. L16:118 weakly suppresses
  24. L10:1473 weakly suppresses

Feeds later layers

  1. L21:812 strongly suppresses
  2. L20:601 strongly suppresses
  3. L19:1237 strongly suppresses
  4. L27:2224 suppresses
  5. L19:1639 excites
  6. L19:277 excites
  7. L24:2266 suppresses
  8. L19:1728 suppresses
  9. L23:1368 suppresses
  10. L19:572 suppresses
  11. L22:2727 suppresses
  12. L19:848 excites
  13. L20:2203 excites
  14. L19:593 suppresses
  15. L20:1710 excites
  16. L19:1070 suppresses
  17. L19:322 excites
  18. L21:325 excites
  19. L20:2652 excites
  20. L27:1818 suppresses
  21. L20:1040 excites
  22. L19:1853 excites
  23. L21:1679 excites
  24. L27:2962 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 -16.92 and -16.392 values between -12.15 and -11.621 values between -11.62 and -11.093 values between -11.09 and -10.563 values between -10.56 and -10.041 values between -10.04 and -9.514 values between -9.51 and -8.983 values between -8.98 and -8.453 values between -8.45 and -7.924 values between -7.92 and -7.392 values between -7.39 and -6.867 values between -6.86 and -6.3311 values between -5.80 and -5.276 values between -5.27 and -4.746 values between -4.74 and -4.217 values between -4.21 and -3.6813 values between -3.68 and -3.1524 values between -3.15 and -2.6231 values between -2.62 and -2.0946 values between -2.09 and -1.5663 values between -1.56 and -1.03108 values between -1.03 and -0.50666 values between -0.50 and 0.031841 values between 0.03 and 0.56980 values between 0.56 and 1.0970 values between 1.09 and 1.621 values between 1.62 and 2.151 values between 3.74 and 4.27offnegativepositiveoften

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. ifferences
  2. t
  3. enuity
  4. dığı
  5. orning
  6. sb
  7. ght
  8. dığımız
  9. entifier
  10. e

Suppresses

  1. Than
  2. -than
  3. _than
  4. 링
  5. Than
  6. board
  7. THAN
  8. 手套
  9. BOARD
  10. icals