Neuron Explorer

Qwen3-0.6B

"from" after change or transfer verbs

A neuron in layer 5 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 lights up on the word "from" when it follows a verb of moving, switching, or changing, such as "switch from," "transform from," or "translate from." Its negative side is very weak and not clearly meaningful. Turning it up did not produce text that clearly showed this pattern.

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

    control circuit continuously monitors the battery's voltage. When the battery is fully charged, the circuit automatically turns on a power inverter and switches the appliance from running on grid power to running on the energy stored in the battery. Then when

  2. #2 peak

    led in a quiet residential neighborhood in Long Beach, Calif., are 4.7 acres of Spanish, Mexican and American history, where families helped transform southern California from its ranching beginnings to a modern, urban society.⏎Rancho Los C

  3. #3 peak

    serious consideration. Now, take a look further for hacks that will turn you from non-native to native English writing prowess.⏎- Like from a speaking perspective This doesn

  4. #4 peak

    28, providing they can switch their children from a 25th percentile teacher to a 75th percentile teacher.⏎

  5. #5 peak

    few seconds:⏎See how “normal” is translated from English to French with more examples in context⏎- Create your own vocabulary list⏎

  6. #6 peak

    of snack manufacturer Snyder’s-Lance Inc. illustrates how this can be applied to a major change initiative. The company’s CEO initiated a major organizational change to move from a more traditional, top-down culture to a servant-leadership culture. A

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: you are being asked to answer a question, and you are trying to figure out what it is. a, i, s, c, o, l, f, m, c, r, d, c, h, r,

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

    temperature)⏎and E is the flux in W/m²⏎More about this formula and background on the material properties in Planck, Stefan

  2. #2 peak

    INTON LOCKHART, Ph. D., LL. D., Author of Laws of

  3. #3 peak

    /or data.⏎Some local statistics:⏎- Smart phone penetration around 51%⏎- >80% use cellphones to communicate⏎- 4G/LTE subscriptions 22%⏎- 4G/LTE coverage 86%⏎

  4. #4 peak

    abyte data transfers every second, initially it is expected the technique will allow data to be stored at a density of 10 petabytes/m2 and using a data transfer rate of 200Gb/s. That’s

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. L3:1858 excites
  2. L3:1907 weakly suppresses
  3. L1:765 weakly excites
  4. L4:1874 weakly suppresses
  5. L1:2454 weakly excites
  6. L4:600 weakly suppresses
  7. L2:1501 weakly suppresses
  8. L3:2458 weakly excites
  9. L2:237 weakly excites
  10. L3:1001 weakly excites
  11. L1:2863 weakly excites
  12. L1:1608 weakly excites
  13. L1:1200 weakly suppresses
  14. L4:2111 weakly suppresses
  15. L3:2149 weakly suppresses "that" after demand or recommend verbs
  16. L4:1635 weakly excites 'light' as an adjective (pale, not heavy)
  17. L4:2649 weakly excites
  18. L0:1270 weakly excites
  19. L1:1480 weakly suppresses
  20. L2:2427 weakly excites
  21. L3:1787 weakly excites
  22. L3:1982 weakly excites
  23. L0:920 weakly suppresses digit '9' inside years and numbers
  24. L1:2393 weakly suppresses

Feeds later layers

  1. L8:2018 excites
  2. L9:229 excites
  3. L18:2272 excites
  4. L14:513 weakly excites
  5. L14:2475 weakly excites
  6. L17:1488 weakly excites
  7. L12:1037 weakly suppresses
  8. L18:1857 weakly suppresses
  9. L13:2762 weakly excites
  10. L17:2683 weakly suppresses
  11. L16:1779 weakly suppresses
  12. L26:1675 weakly suppresses
  13. L16:1514 weakly suppresses
  14. L22:2582 weakly excites
  15. L24:1070 weakly excites
  16. L6:1168 weakly excites
  17. L6:364 weakly excites
  18. L7:2179 weakly suppresses
  19. L23:2868 weakly suppresses
  20. L25:454 weakly suppresses
  21. L9:723 weakly suppresses
  22. L23:1660 weakly excites
  23. L24:2083 weakly suppresses
  24. L27:1717 weakly suppresses 'st' starting a word fragment

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.

3 values between -0.67 and -0.584 values between -0.58 and -0.508 values between -0.50 and -0.4131 values between -0.41 and -0.32103 values between -0.32 and -0.24288 values between -0.24 and -0.15739 values between -0.15 and -0.061473 values between -0.06 and 0.03822 values between 0.03 and 0.11291 values between 0.11 and 0.2089 values between 0.20 and 0.2924 values between 0.29 and 0.3710 values between 0.37 and 0.461 values between 0.46 and 0.554 values between 0.55 and 0.644 values between 0.64 and 0.721 values between 0.72 and 0.811 values between 0.81 and 0.902 values between 0.90 and 0.981 values between 0.98 and 1.072 values between 1.16 and 1.251 values between 1.25 and 1.331 values between 1.59 and 1.681 values between 1.68 and 1.771 values between 1.86 and 1.941 values between 2.03 and 2.121 values between 2.29 and 2.381 values between 2.73 and 2.81offnegativepositiveoften

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. 换了
  2. /question
  3. iated
  4. aque
  5. rawn
  6. ftware
  7. ardi
  8. 腑
  9. hdr
  10. 荷

Suppresses

  1. )[-
  2. pop
  3. отов
  4. IMPLIED
  5. inness
  6. pop
  7. еть
  8. etimes
  9. CrossRef
  10. houses