Neuron Explorer

Qwen3-0.6B

'to' after verbs of change

A neuron in layer 16 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 'to' when it follows a verb of change or movement and points to where something ends up, like 'warmed up to 3.5 degrees' or 'turn orange to red'. When it was turned up, the model's output got stuck repeating 'to to' and similar phrases.

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

    racial equality. Indeed, it was the amendment process that created the legal basis to abolish slavery and assure equal political rights for black Americans. As the fight goes on to assert and realize the Constitutional promise of equal representation for all, we humbly

  2. #2 peak

    flipped and rotated in all directions. This may help to make some furtive words more readable. Also the spacing, boldness and capital case can be switched to make the letters stand out from their context. Optionally a handy hover-cross is shown

  3. #3 peak

    significant part of Alaska is spared.⏎Permafrost is underlying about 85% of Alaska. Over the past several decades, permafrost has warmed up to 3.5 degrees Fahrenheit and Alaska’s permafrost is now essentially melting

  4. #4 peak

    izing the aging elementary facilities in its portfolio. The first school being built with net-zero principles is Douglas MacArthur Elementary which first opened in 1943 to serve families who came to the greater

  5. #5 peak

    2030 that is envisaged in the Sustainable Development Scenario. The buildings sector also remains off track, with emissions rising again in 2019 to an all-time high. Another concerning development was the slowdown in fuel economy improvements around

  6. #6 peak

    Conditions affecting the tonsils and adenoids⏎Several conditions may affect the tonsils and adenoids in a child:⏎Subscribe to our mailing list to get the updates to your email inbox...<|endoftext|>Detoxification - alcohol; Detox - alcohol⏎

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 have to, from to, and vice a versa, and vice a versa. No, no, no. ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...

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

    );⏎void deallocate_model(void *);⏎And that is why I am a header Nazi. Here's how I do<|endoftext|>When you work with people, conflict is bound to happen. The dictionary defines conflict as, “A serious disagreement or argument, typically

  2. #2 peak

    problem arises when all or most of the benefits of any pollution-emitting activity accrue to one- nation, while all or most of the costs of that activity accrue

  3. #3 peak

    uler vs. Pie Concept.⏎Webster defines Truth as Reality. But reality to whom? (What does this have to do with consumer research? We will get to that later.) We have often heard the saying that, "Truth is in the

  4. #4 peak

    988413967-1-768x146-1-300x57.png 0 0 admin https://

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. L4:328 suppresses 'to' after 'so as' and infinitive cues
  2. L14:392 suppresses
  3. L1:2296 excites
  4. L3:1152 suppresses
  5. L5:3009 excites 'to' after a verb or noun
  6. L5:2251 suppresses
  7. L1:906 suppresses
  8. L15:78 excites
  9. L10:43 suppresses "terms" in "in terms of"
  10. L6:2064 weakly suppresses
  11. L2:1585 weakly excites
  12. L6:1534 weakly suppresses
  13. L4:221 weakly suppresses
  14. L10:445 weakly excites
  15. L1:1464 weakly excites
  16. L4:2047 weakly suppresses
  17. L0:1551 weakly suppresses
  18. L10:132 weakly excites
  19. L6:2543 weakly excites
  20. L4:862 weakly suppresses 'to' after 'too big' or 'right angles'
  21. L1:146 weakly excites
  22. L5:418 weakly suppresses
  23. L6:211 weakly excites
  24. L0:178 weakly suppresses

Feeds later layers

  1. L17:2821 strongly excites
  2. L17:1297 excites
  3. L19:1874 excites
  4. L18:1408 excites
  5. L18:1001 excites
  6. L23:373 excites 'to' after a noun or adjective
  7. L19:858 excites
  8. L17:2527 excites
  9. L21:2012 excites "to" before an infinitive or noun
  10. L18:2760 excites 'to' after a verb of sharing or using
  11. L17:1843 excites
  12. L25:286 excites
  13. L17:472 excites
  14. L21:1657 excites
  15. L18:285 suppresses
  16. L21:1081 excites
  17. L18:90 excites
  18. L17:303 excites
  19. L19:1237 suppresses
  20. L17:1884 suppresses
  21. L18:1675 weakly excites
  22. L18:176 weakly excites
  23. L19:1463 weakly suppresses
  24. L17:1122 weakly excites

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.

40 values between -1.36 and -0.92796 values between -0.92 and -0.482542 values between -0.48 and -0.03382 values between -0.03 and 0.4136 values between 0.41 and 0.8618 values between 0.86 and 1.3012 values between 1.30 and 1.749 values between 1.74 and 2.198 values between 2.19 and 2.636 values between 2.63 and 3.082 values between 3.08 and 3.526 values between 3.52 and 3.964 values between 3.96 and 4.416 values between 4.41 and 4.854 values between 4.85 and 5.293 values between 5.29 and 5.743 values between 5.74 and 6.184 values between 6.18 and 6.632 values between 6.63 and 7.071 values between 7.07 and 7.511 values between 7.51 and 7.963 values between 8.40 and 8.852 values between 8.85 and 9.293 values between 9.29 and 9.731 values between 9.73 and 10.182 values between 10.18 and 10.622 values between 10.62 and 11.062 values between 11.06 and 11.511 values between 11.51 and 11.951 values between 12.40 and 12.841 values between 13.73 and 14.173 values between 14.17 and 14.621 values between 14.62 and 15.061 values between 15.95 and 16.39offnegativepositiveoften

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. /from
  2. 的就是
  3. (screen
  4. gether
  5. yal
  6. ulet
  7. render
  8. ombat
  9. validate
  10. izi

Suppresses

  1. ARED
  2. heit
  3. IFIED
  4. lation
  5. ified
  6. lijk
  7. roleum
  8. 扫一扫
  9. ification
  10. -fat