Neuron Explorer

Qwen3-0.6B

'for' after an award or honour

A neuron in layer 18 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 strongly on the word 'for' when it introduces the reason someone was honoured, rewarded or punished, like 'won the Nobel Peace Prize for' or 'an award for'. Turned up, the model fell into repeated 'for the community' phrases.

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

    , Denison Miller was reappointed from 1 June 1919 for a further seven years. He was knighted in 1920 for services to<|endoftext|>But what exactly is Diabetes? There are a lot of myths and

  2. #2 peak

    she is the author of Bulldozer: Demolition and Clearance of the Postwar Landscape, winner of the 2017 Lewis Mumford Prize for the best book in American planning history. She is particularly interested in post-World

  3. #3 peak

    website, Equipe Cousteau , and the educational and scientific websites of the International Polar Foundation. In February 2009 CAML received an award for 'overall exellence' from its parent program, the Census of Marine Life

  4. #4 peak

    family was strongly loyal to the Confederate cause. Her older sister Eugenia married a U.S. congressman from Alabama, and was herself imprisoned several times by Federal authorities for her Southern nationalism. A younger brother, Samuel, was a major in the Georgia

  5. #5 peak

    the past⏎Before he became president of Poland in 1990, Lech Walesa had already won the 1983 Nobel Peace Prize for his dangerous human rights fight in the face of a communist government. His work eventually

  6. #6 peak

    numerous awards, including being named best business start-up in the Institution of Chemical Engineers Global Awards 2020 and being highly commended at the same event for energy saving.⏎Dr Feng and his team also have a start-up and are working

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 idea that a person can be in a state of [fill in the blank] [1] ] ] is not just a temporary, like a bird can be in a state of [fill in the] ]

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

    , you need to do it soon. Failure to do so is a felony punishable by up to five years in prison or a $250,000 fine.⏎And the system has not been updated since women were allowed to serve in

  2. #2 peak

    , but also that it can monitor,<|endoftext|>This material must not be used for commercial purposes, or in any hospital or medical facility. Failure to comply may result in legal action.⏎WHAT YOU NEED TO KNOW:⏎Diphtheria is a disease

  3. #3 peak

    take in the privacy of our own homes, making it particularly convenient in our busy day-to-day lives.⏎This post may contain affiliate links. I will earn a small commission, at no extra cost to you, if you make a purchase through an

  4. #4 peak

    oil spilling should be fined and the laws should be tightened. I also wish the world would always be in peace

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. L13:214 excites
  2. L12:2776 suppresses
  3. L10:2857 excites
  4. L14:2332 suppresses
  5. L15:2769 excites
  6. L16:2910 excites
  7. L11:1279 suppresses
  8. L12:468 suppresses
  9. L9:2582 excites
  10. L9:2099 excites
  11. L9:2435 suppresses
  12. L13:2329 excites
  13. L6:2553 excites 'for' after praise words like 'known for'
  14. L6:12 excites
  15. L8:1588 weakly suppresses
  16. L15:1804 weakly excites
  17. L6:617 weakly excites
  18. L7:1731 weakly excites
  19. L14:2607 weakly excites
  20. L13:378 weakly excites
  21. L17:210 weakly suppresses
  22. L7:1558 weakly suppresses
  23. L12:2852 weakly excites
  24. L9:2823 weakly excites

Feeds later layers

  1. L21:1609 strongly excites
  2. L22:2351 excites
  3. L21:2230 weakly excites
  4. L19:1796 weakly excites
  5. L21:1435 weakly excites
  6. L20:330 weakly excites
  7. L21:2368 weakly excites
  8. L19:931 weakly suppresses
  9. L20:2367 weakly excites
  10. L19:2117 weakly suppresses
  11. L23:2460 weakly excites
  12. L19:3050 weakly excites
  13. L23:1038 weakly excites
  14. L19:747 weakly excites
  15. L24:1162 weakly excites
  16. L26:82 weakly suppresses
  17. L26:1064 weakly excites
  18. L21:2729 weakly excites
  19. L23:553 weakly suppresses
  20. L20:1853 weakly excites
  21. L26:2970 weakly excites
  22. L20:2725 weakly suppresses
  23. L20:1804 weakly suppresses
  24. L19:757 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.

1 values between -9.16 and -8.021 values between -3.48 and -2.3442 values between -2.34 and -1.212990 values between -1.21 and -0.07800 values between -0.07 and 1.0738 values between 1.07 and 2.2012 values between 2.20 and 3.3412 values between 3.34 and 4.484 values between 4.48 and 5.614 values between 5.61 and 6.751 values between 6.75 and 7.882 values between 7.88 and 9.021 values between 35.15 and 36.28offnegativepositiveoften

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. ��
  2. 功劳
  3. pons
  4. 行为
  5. haber
  6. positor
  7. ButtonItem
  8. CLU
  9. dish
  10. adv

Suppresses

  1. 限り
  2. enda
  3. arat
  4. ilogue
  5. われ
  6. relude
  7. hope
  8. ounter
  9. '%'
  10. 感覺