Neuron Explorer

Qwen3-0.6B

gravity word pieces (gravit-)

A neuron in layer 2 of Qwen3-0.6B. Its strongest snippets are mixed: they peak on different tokens. It is featured, so it has steering outputs.

What this neuron does

This neuron reacts to the middle of words built on 'gravit' or 'grav', such as 'Gravitational waves', 'gravitationally bound' and 'gravity' as split into pieces. It is also weakly tied to 'gravis' in myasthenia gravis. When it was turned up, the model's text did not clearly move toward physics or gravity.

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

    their existence a century ago, physicists have been on the hunt for gravitational waves, ripples in the fabric of spacetime. That hunt is now over. Gravitational waves exist, and we’ve found them.⏎That’s according to researchers at

  2. #2 peak

    etime ripples started to get serious.⏎But the first generation LIGO experiment, which ran for eight years, wasn’t sensitive enough. Which is understandable. Gravitational waves are minuscule — the atomic jitters that pass through our world

  3. #3 peak

    about three times the mass of the sun was converted to energy in a fraction of a second.⏎The discovery has been accepted for publication in Physical Review Letters.⏎Gravitational waves are ripples in the universe caused by some of the most energetic cosmic

  4. #4 peak

    by Gizmodo have been flying for a month, the LIGO team wanted to be absolutely certain before making an official announcement.⏎That announcement has just come. Gravitational waves were observed on 14 September 2015, at

  5. #5 peak

    for advanced students and researchers.⏎Catalogue of Galactic Supernova Remnants (SNRs)⏎Listing of over 200 supernova remnants.⏎A Gravitationally Powered Oscillatory Pulsar Model⏎Proposed advanced theory of mainstream

  6. #6 peak

    which the Milky Way and Andromeda are the dominant members. The rest of the galaxies are mostly small satellites known as "dwarf galaxies" that are gravitationally bound to these two galaxies. The Large and Small Magellanic Clouds

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: "I am a senior executive in the company, and I have been asked to join the executive of the European Commission of Consumer Reports, and I have been told to the European Commission of Consumer Reports, and I have been told to the European

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

    architecture and evolved from earlier Romanesque buildings. Gothic architecture combines elements of art and buildings, theology and Gothic literature. Gothic architecture has become typified by cathedrals, fortresses, palaces and churches.⏎thai interior Gothic architecture

  2. #2 peak

    ched structure from the Gothic church is characteristic of this specific style of architecture. Gothic cathedrals were typically designed with an individual wall, which has been both decorated or

  3. #3 peak

    shrines, around which cities grew. Ancient cities usually had large temple precincts with monumental religious buildings. Many medieval cities were built near monasteries and cathedrals

  4. #4 peak

    , Early English, Early English north and south transepts, Eboracum, England, Europe, Freemasonry, glass conservation, Gothic, Gothic cathedrals, Great Britain, Great East Window, Heritage Lottery Fund, hugely significant

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. L0:1224 suppresses
  2. L0:1301 suppresses
  3. L0:178 suppresses
  4. L1:5 weakly excites
  5. L1:340 weakly excites
  6. L0:1388 weakly excites
  7. L1:3038 weakly excites
  8. L1:1910 weakly excites
  9. L1:178 weakly excites
  10. L1:2891 weakly excites
  11. L1:1545 weakly suppresses
  12. L1:1857 weakly suppresses
  13. L0:1484 weakly excites
  14. L0:2601 weakly suppresses
  15. L1:2318 weakly suppresses
  16. L1:2381 weakly suppresses
  17. L0:2206 weakly suppresses
  18. L1:486 weakly suppresses
  19. L0:1699 weakly excites
  20. L1:1110 weakly suppresses
  21. L1:1973 weakly suppresses
  22. L1:2768 weakly suppresses
  23. L1:905 weakly excites
  24. L0:949 weakly excites

Feeds later layers

  1. L4:743 excites
  2. L3:2925 weakly suppresses
  3. L3:1528 weakly suppresses
  4. L20:1656 weakly suppresses
  5. L20:2034 weakly suppresses
  6. L8:2186 weakly excites
  7. L7:1307 weakly suppresses
  8. L24:2707 weakly suppresses
  9. L21:1933 weakly suppresses
  10. L3:2223 weakly suppresses
  11. L12:3040 weakly excites
  12. L3:1318 weakly suppresses
  13. L16:2712 weakly excites
  14. L13:1033 weakly suppresses
  15. L19:1466 weakly excites
  16. L6:1952 weakly excites
  17. L8:2850 weakly excites
  18. L9:1182 weakly excites
  19. L4:2388 weakly suppresses
  20. L3:2663 weakly excites
  21. L16:2889 weakly suppresses
  22. L22:591 weakly suppresses
  23. L9:2450 weakly suppresses
  24. L13:2380 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 -2.43 and -2.261 values between -2.26 and -2.091 values between -0.74 and -0.572 values between -0.57 and -0.4046 values between -0.40 and -0.23695 values between -0.23 and -0.062782 values between -0.06 and 0.11354 values between 0.11 and 0.2819 values between 0.28 and 0.454 values between 0.45 and 0.611 values between 0.61 and 0.781 values between 0.78 and 0.951 values between 4.17 and 4.34offnegativepositiveoften

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. 提款
  2. atas
  3. Specification
  4. 擒
  5. 遗憾
  6. stopping
  7. ))[
  8. ちょう
  9. aho
  10. ])[

Suppresses

  1. erville
  2. Nacht
  3. بيض
  4. gree
  5. paque
  6. irteen
  7. izzes
  8. ertain
  9. izzle
  10. ряд