Neuron Explorer

Qwen3-0.6B

the word "the" starting a text

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 most strongly to the word "The" when it is the first word of a new piece of text, especially right after an end-of-document marker. Its negative side is weak and hard to read. Turning it up or down did not produce a clear change in theme.

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

    <|endoftext|>The Battle of the Kasserine Pass was a battle that took place during the Tunisia

  2. #2 peak

    The quest to achieve the impossible fuels creativity, spawns new fields of inquiry, illuminates

  3. #3 peak

    The studio probably had little choice but to create a form letter.⏎I don’t think

  4. #4 peak

    The recipient previously opened the message⏎Still, also the recall won’t work If a

  5. #5 peak

    The Negro Speaks of Rivers,” written the summer after he graduated from high school in

  6. #6 peak

    The unused funds — called Universal Service and Access Funds (USAFs) — are

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: the number of the number of the number of the number of the number of the number of the number of the number of the number of the number of the number of the number of the number of the number of the number of the number of

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

    lots there can be lots of traffic. The noise that is created during the drilling operation affects wildlife. Their migration routes may be affected.⏎Drilling operations need heavy equipment like bulldozers. These may affect the vegetation and leave permanent damage on the

  2. #2 peak

    other school supplies. They are also used to carry lunch and other items.⏎History of backpacks⏎Originally designed to carry heavy equipment, korean backpack for school became popular for carrying books in the early 1

  3. #3 peak

    tabs are an important statistical process that summarizes two or more variables of categorical data into a contingency table. They

  4. #4 peak

    tasks that require operating electrical equipment.⏎This becomes even more common for people who work exclusively in the electrical department, fixing up wiring issues, using different highly-charged equipment, etc.

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. L1:2703 weakly excites
  2. L11:2732 weakly suppresses
  3. L2:55 weakly excites
  4. L2:321 weakly excites
  5. L16:2769 weakly suppresses
  6. L1:997 weakly suppresses
  7. L26:614 weakly excites
  8. L6:486 weakly suppresses
  9. L27:490 weakly suppresses
  10. L20:1844 weakly suppresses
  11. L3:1118 weakly suppresses
  12. L10:999 weakly suppresses
  13. L21:342 weakly excites
  14. L22:1669 weakly suppresses
  15. L11:428 weakly excites
  16. L1:1262 weakly excites
  17. L1:1321 weakly excites
  18. L8:281 weakly excites
  19. L12:1783 weakly suppresses
  20. L3:236 weakly suppresses
  21. L11:97 weakly excites
  22. L10:408 weakly suppresses
  23. L9:633 weakly excites
  24. L26:1514 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.

26 values between -0.25 and -0.1952 values between -0.19 and -0.14281 values between -0.14 and -0.09764 values between -0.09 and -0.031343 values between -0.03 and 0.021017 values between 0.02 and 0.07354 values between 0.07 and 0.1356 values between 0.13 and 0.189 values between 0.18 and 0.233 values between 0.23 and 0.281 values between 0.34 and 0.391 values between 0.55 and 0.601 values between 1.82 and 1.87offnegativepositiveoften

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. .toList
  2. além
  3. ToFront
  4. orraine
  5. istics
  6. 请问
  7. /*#__
  8. ensis
  9. 并对
  10. bidden

Suppresses

  1. shore
  2. че
  3. ยะ
  4. Placement
  5. cit
  6. 人心
  7. erin
  8. Frameworks
  9. hari
  10. umerator