Neuron Explorer

Qwen3-0.6B

line break after a title

A neuron in layer 0 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 reacts to the line break right after a short title or heading, just before the body text starts. Turning it up made the model's writing drift into unrelated topics, with no clear link to the trigger.

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

    essayist,<|endoftext|>How to Prevent Dry Eyes During Summer Months⏎Summer is around the corner, and one can feel the heat of the scorch

  2. #2 peak

    . These foods include dairy products, legumes and grains.⏎How Much Protein On Paleo Diet⏎The goal of the paleo diet is to return to a way of eating that

  3. #3 peak

    <|endoftext|>Genetic Testing Resources⏎Inherited retinal degenerative diseases, like retinitis pigmentosa or St

  4. #4 peak

    Manipulation⏎- Spirit Physiology⏎- Undead Physiology⏎- Despite being able to survive without

  5. #5 peak

    Initiative Prevention Curriculum⏎Target Audience: Middle school-aged youth⏎Administration: This curriculum can be used by

  6. #6 peak

    ale Revolution and Climate Change⏎- The Shale Revolution and Earthquakes⏎- The Economic Impacts of

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 will not be able to do that, I will not be able to do that. I will not be able to do that. I will not be able to do that. I will not be able to do that. I will not

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

    inium which is fairly easy to identify and is blooming now.⏎Deerberry – Vaccinium stamineum – has relatively large, attractive, abundant, white bell-shaped flowers

  2. #2 peak

    that identification is difficult and natural hybrids are common. Further, huckleberries, which are now in the genus Gaylussacia were once in the genus Vaccinium. And one of the Vacciniums I think I may have identified here goes

  3. #3 peak

    ").⏎In 1608 tried to patent a telescope fitted with a pair of sunglasses, a concave and another convex, which it termed perspicillum: was born the first astronomical telescope refractor. The manufacturer of lenses designed it with

  4. #4 peak

    're saving.⏎For More Information...⏎Phone: 1-888-STAR-YES (1-8

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

None: this is the first layer.

Feeds later layers

  1. L1:272 suppresses
  2. L14:387 suppresses
  3. L22:2375 suppresses
  4. L18:1370 weakly excites
  5. L2:2713 weakly suppresses
  6. L10:783 weakly suppresses
  7. L16:84 weakly excites
  8. L14:2319 weakly excites
  9. L16:2546 weakly excites
  10. L1:655 weakly suppresses
  11. L24:1117 weakly excites
  12. L14:2029 weakly excites
  13. L3:2516 weakly suppresses
  14. L14:371 weakly suppresses
  15. L2:368 weakly suppresses
  16. L8:2337 weakly suppresses
  17. L24:1084 weakly suppresses
  18. L2:2852 weakly suppresses
  19. L13:753 weakly suppresses
  20. L1:918 weakly suppresses
  21. L27:2560 weakly suppresses
  22. L13:152 weakly excites
  23. L1:138 weakly suppresses
  24. L2:1102 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 -0.40 and -0.361 values between -0.36 and -0.324 values between -0.32 and -0.272 values between -0.27 and -0.2312 values between -0.23 and -0.1931 values between -0.19 and -0.1594 values between -0.15 and -0.11240 values between -0.11 and -0.06577 values between -0.06 and -0.021133 values between -0.02 and 0.02827 values between 0.02 and 0.06525 values between 0.06 and 0.11290 values between 0.11 and 0.15105 values between 0.15 and 0.1934 values between 0.19 and 0.236 values between 0.23 and 0.277 values between 0.27 and 0.323 values between 0.32 and 0.363 values between 0.36 and 0.401 values between 0.40 and 0.441 values between 0.44 and 0.482 values between 0.53 and 0.571 values between 0.65 and 0.702 values between 0.70 and 0.742 values between 0.78 and 0.821 values between 0.91 and 0.951 values between 0.99 and 1.031 values between 1.07 and 1.121 values between 1.24 and 1.28offnegativepositiveoften

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. mıştı
  2. ctors
  3. �
  4. 十条
  5. erner
  6. izedName
  7. 韬
  8. estad
  9. ertools
  10. 中华文化

Suppresses

  1. лина
  2. 捡
  3. imit
  4. 凇
  5. 彗
  6. Sprites
  7. }:
  8. cones
  9. avit
  10. ints