Neuron Explorer

Qwen3-0.6B

end of document boundary tokens

A neuron in layer 13 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 mostly lights up on the marker that signals the end of one document and the start of an unrelated one, so it seems to notice where a new piece of text begins. Turning it up made the model's writing wander into unrelated topics, with no clear theme.

Written by Claude from this neuron's activations and steering outputs. Confidence: low. 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

    code<|endoftext|>Jet engines, or gas turbines, suck in air at the front with a fan

  2. #2 peak

    are found, there is no way<|endoftext|>We won’t stop until we’ve eradicated HIV/AIDS⏎The Drug Development Process

  3. #3 peak

    can we say we believe in it? It is up to me and you.”⏎Adafruit publishes a wide range of writing and video content, including interviews and reporting on the maker market and the wider technology world. Our standards page is intended as a

  4. #4 peak

    groups<|endoftext|>What happens when you mix vinegar and baking powder?⏎When baking soda is mixed with

  5. #5 peak

    | Volume 25, Issue 1⏎That was the end of the campaigning in 1776. Washington went

  6. #6 peak

    the Students from Student table whose marks is more than Average of all marks.⏎(Select Avg(M<|endoftext|>Euler's Equation and the Physical Reality. =. a) Euler's Equation

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 think it is possible to find a set of points, the problem is that it's hard to me. i think it's not hard, but i can't think it's not. it's just think it's not... i can

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

    % off) $17.25 new (22% off) $21.95 direct from Amazon Amazon page|⏎Setup an account with your affiliations in order to access resources via your University's proxy server⏎Configure custom

  2. #2 peak

    conjugate, write the correct conjugate acid or base for the species in the pair.⏎a. H2O, OH-⏎Frequently Asked Questions⏎What scientific concept do you need to know in order to solve this

  3. #3 peak

    quality lesson planning resources, fast!⏎Share & remix collections to collaborate.⏎Organize your curriculum with collections. Easy!⏎Have time to be more creative & energetic with your students!⏎In this verb tense review worksheet, students are given sentences and must find

  4. #4 peak

    collections to collaborate.⏎Organize your curriculum with collections. Easy!⏎Have time to be more creative & energetic with your students!⏎Students explore the Orange Revolution. In this Orange Revolution lesson, students examine the climate of the Ukrainian protest and investigate the role

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. L5:203 suppresses
  2. L3:70 suppresses
  3. L10:816 excites
  4. L5:2007 suppresses
  5. L10:1580 suppresses
  6. L6:2030 suppresses
  7. L10:84 suppresses
  8. L1:816 weakly excites
  9. L9:1688 weakly excites
  10. L10:632 weakly excites
  11. L3:1257 weakly excites
  12. L7:142 weakly excites
  13. L12:1186 weakly suppresses
  14. L2:1031 weakly excites
  15. L8:32 weakly excites
  16. L5:619 weakly excites
  17. L5:238 weakly excites
  18. L6:520 weakly excites
  19. L11:1629 weakly excites
  20. L0:1650 weakly excites
  21. L6:216 weakly excites
  22. L8:637 weakly excites
  23. L5:2738 weakly suppresses
  24. L1:179 weakly suppresses

Feeds later layers

  1. L19:1479 excites
  2. L15:885 excites
  3. L18:2202 excites
  4. L14:886 excites
  5. L17:1953 excites
  6. L19:441 excites
  7. L15:344 excites
  8. L15:2717 suppresses
  9. L14:1274 excites
  10. L20:1081 suppresses
  11. L20:2120 suppresses "so" opening a purpose clause
  12. L19:2232 weakly excites end-of-document boundary token
  13. L19:624 weakly excites
  14. L17:647 weakly suppresses
  15. L17:2481 weakly suppresses
  16. L25:630 weakly suppresses
  17. L22:2221 weakly suppresses
  18. L14:901 weakly excites mid-word fragments in technical terms
  19. L26:2340 weakly suppresses
  20. L26:2358 weakly suppresses
  21. L14:56 weakly excites document ends and list line breaks
  22. L19:1083 weakly suppresses
  23. L16:49 weakly suppresses
  24. L23:2124 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.

4 values between -0.57 and -0.5211 values between -0.52 and -0.4719 values between -0.47 and -0.4247 values between -0.42 and -0.3792 values between -0.37 and -0.32163 values between -0.32 and -0.26225 values between -0.26 and -0.21356 values between -0.21 and -0.16511 values between -0.16 and -0.11521 values between -0.11 and -0.06596 values between -0.06 and -0.01514 values between -0.01 and 0.04353 values between 0.04 and 0.10216 values between 0.10 and 0.15136 values between 0.15 and 0.2067 values between 0.20 and 0.2533 values between 0.25 and 0.3021 values between 0.30 and 0.3514 values between 0.35 and 0.404 values between 0.40 and 0.461 values between 0.46 and 0.511 values between 0.66 and 0.711 values between 0.71 and 0.761 values between 0.82 and 0.871 values between 1.43 and 1.48offnegativepositiveoften

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. alm
  2. /*⏎⏎
  3. dirig
  4. [keys
  5. görm
  6. icultural
  7. mitter
  8. -root
  9. mart
  10. 朵

Suppresses

  1. ughter
  2. `]
  3. "], ⏎
  4. :],
  5. §Ã
  6. sẵn
  7. もなく
  8. -dess
  9. "]],⏎
  10. "],