Neuron Explorer

Qwen3-0.6B

word after sentence-initial 'You'

A neuron in layer 0 of Qwen3-0.6B. Its strongest snippets are fairly consistent: several of them peak on the same token. It is featured, so it has steering outputs.

What this neuron does

This neuron fires on the word that comes right after 'You' at the start of a sentence, like 'know' in 'You know' or 'can' in 'You can'. It is most common in friendly advice and conversational text. Turning it up or down did not clearly change the model's writing.

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

    make sure they will know what steps to take, and don’t waste valuable time spinning their wheels.⏎Does your teen have the “planning gene”?⏎You know when you see them – paper planners, calendars on the phone, everything laid out

  2. #2 peak

    40, and others recommend starting at age 50. In general, all men should create a proactive prostate health plan –in cooperation with their trusted Urologist- that is right for them based on their lifestyle and family history.

  3. #3 peak

    STEM paintbrush engineering challenge!⏎Put together a quick rubber or loom band busy bag to take with you on travels or elsewhere and keep your toddler entertained.⏎You know how geoboards are useful for strengthening little fingers? Well, they will surely

  4. #4 peak

    English Language, 4th Edition⏎- adj. Ending in a peak; pointed: a peaked cap.⏎- adj. Having a sickly appearance: You're looking a little peaked today.⏎from Wiktionary, Creative Commons Attribution/Share

  5. #5 peak

    of articulated skeletal elements, fossil remains with some or all skeletal elements found in the same spatial relation to one another that they would have had in life. [You know. With the knee bone connected to the thigh bone, etc.]⏎|Arch

  6. #6 peak

    believe that we owe this story to those who came before us.⏎As a young person growing up in this country, I remember my parents telling me, “You know what?⏎We have to get out of this country and go to another country,

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 want to be able to know the information of the other people and their experiences. So I want to know the experiences of others, and I want to be able to know their information. So I want to be able to know their information

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

    different ways that coffee can be brewed.⏎There are two main types of coffee plants. The Coffea arabica, the most common. Most

  2. #2 peak

    .⏎You can also purchase the Almanac Monthly at the iTunes Store !<|endoftext|>Learning in diss

  3. #3 peak

    the lawn, using organic fertilizers, aerating to reduce soil compaction, and de-thatching, to make your lawn healthier and more vigorous naturally.⏎You can improve the health of you soil. Common weeds considered by many as pests are a

  4. #4 peak

    of cinnamon trees. There are two main types of cinnamon: Cassia and Ceylon.⏎Ceylon cinnamon comes

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. L16:1372 weakly excites
  2. L22:1594 weakly excites
  3. L12:1473 weakly suppresses
  4. L12:2908 weakly excites
  5. L17:1979 weakly suppresses
  6. L5:334 weakly excites
  7. L6:158 weakly excites
  8. L10:2688 weakly excites
  9. L5:594 weakly excites
  10. L16:1828 weakly excites
  11. L16:2383 weakly excites
  12. L12:2161 weakly excites
  13. L12:2604 weakly excites
  14. L25:2847 weakly suppresses
  15. L11:685 weakly suppresses
  16. L9:651 weakly suppresses
  17. L6:2158 weakly excites
  18. L21:524 weakly excites
  19. L27:2222 weakly excites
  20. L4:1255 weakly suppresses
  21. L13:1754 weakly excites
  22. L5:803 weakly excites
  23. L5:2408 weakly excites
  24. L21:1727 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.

2 values between -1.37 and -1.301 values between -1.06 and -0.981 values between -0.83 and -0.751 values between -0.75 and -0.671 values between -0.67 and -0.591 values between -0.51 and -0.431 values between -0.43 and -0.351 values between -0.35 and -0.2822 values between -0.28 and -0.2096 values between -0.20 and -0.12541 values between -0.12 and -0.042620 values between -0.04 and 0.04517 values between 0.04 and 0.1271 values between 0.12 and 0.2016 values between 0.20 and 0.276 values between 0.27 and 0.354 values between 0.35 and 0.433 values between 0.43 and 0.511 values between 0.59 and 0.671 values between 1.22 and 1.291 values between 1.69 and 1.77offnegativepositiveoften

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. etics
  2. �
  3. dings
  4. 遇到了
  5. self
  6. urations
  7. cher
  8. idence
  9. لاق
  10. PAGE

Suppresses

  1. StackSize
  2. ritt
  3. FML
  4. ضح
  5. ijkl
  6. .observable
  7. ligt
  8. 榜
  9. instrumental
  10. Flint