Neuron Explorer

Qwen3-0.6B

'that' after 'the way' or 'the first time'

A neuron in layer 4 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 lights up on the word 'that' when it follows phrases like 'the way' or 'the first time', as in 'the way that decimals are written'. Turning it up did not produce a clear pattern in the continuations.

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

    before the answer.⏎Subtracting money is easy, too, so long as it is once more remembered that money has to be written and subtracted the way that decimals are written and subtracted. Dollars go to the left of the decimal point

  2. #2 peak

    happy to report that the seeds that were transported to the moon have sprouted! While plants have grown on the International Space Station before, this is the first time that any biological material has grown on the moon. Why is this so important? Because

  3. #3 peak

    3 August 2012⏎Have you ever stepped out into a narrow city street and began to cough? You may have been a victim of the way that 'street canyons' trap polluted air, keeping harmful substances like nitrogen dioxide

  4. #4 peak

    study was posted online Dec. 29 in advance of print publication in the journal Proceedings of the National Academy of Sciences.⏎While this is not the first time that microbes in the human intestinal tract have been sequenced as a community, this is

  5. #5 peak

    should be able to find references or spell the word the way that it sounds.⏎Students should be writing for a variety of purposes.<|endoftext|>The Solar

  6. #6 peak

    harmful ones are bred out. But the “adaptive” nature of the changes natural selection (praise be upon it) wreaks may not be the only way that complexity grew.⏎Protein flaws called “dehydrons” may have made proteins

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: what is the value of the word, what is the value of the word, what is the value of the word, what is the value of the word, what is the value of the word, what is the value of the word,

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

    first author, Dr Damian Scarf of the University of Otago’s Department of Psychology, they performed this feat by tracking the statistical likelihood that “bigrams,” letter pairs such as “EN” and “AL,” were more likely associated with

  2. #2 peak

    - The messages contained within⏎- Illustrates human behaviour in any given situation and/or predicament⏎- Gives rules on life and living⏎- The romance of the narrative⏎- You have heroes

  3. #3 peak

    itself because its already in your mind, because if it does completely undo itself then there’s no memory of that experience anymore. People need to keep these memories because; these memories or remembered experiences make you wise and give you knowledge. Either congenial

  4. #4 peak

    ]⏎- In case of other than assigned generator class, suppose if our generator class name is Increment means hibernate it self will assign the primary key id value into the database right [ other than assigned generator, hibernate only used to take care the

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. L3:115 suppresses
  2. L3:2369 weakly suppresses
  3. L2:111 weakly suppresses
  4. L3:50 weakly excites
  5. L1:2493 weakly excites
  6. L3:186 weakly suppresses
  7. L1:1464 weakly excites
  8. L1:1064 weakly suppresses
  9. L3:1359 weakly suppresses
  10. L2:2349 weakly suppresses word pieces inside cinnamon, China, cabar
  11. L3:421 weakly suppresses
  12. L3:2858 weakly suppresses
  13. L2:704 weakly excites
  14. L0:1925 weakly suppresses
  15. L0:2903 weakly suppresses
  16. L3:691 weakly excites
  17. L1:2503 weakly suppresses
  18. L0:404 weakly excites
  19. L0:2414 weakly excites
  20. L2:2132 weakly excites
  21. L1:2990 weakly excites
  22. L0:1271 weakly excites
  23. L3:187 weakly excites
  24. L1:2358 weakly suppresses

Feeds later layers

  1. L7:114 excites
  2. L7:681 excites 'from' after prevent or stop
  3. L13:1439 excites end of document token
  4. L6:689 excites
  5. L5:1909 excites
  6. L8:392 excites
  7. L15:638 weakly excites "the" inside institution names
  8. L11:1272 weakly suppresses unclear: sentence ends in factual prose
  9. L14:1200 weakly excites
  10. L8:81 weakly excites
  11. L17:181 weakly excites
  12. L16:269 weakly suppresses
  13. L9:2354 weakly excites
  14. L27:2666 weakly suppresses
  15. L15:1554 weakly excites
  16. L17:359 weakly excites
  17. L13:929 weakly suppresses
  18. L17:1894 weakly suppresses
  19. L6:1524 weakly excites
  20. L6:244 weakly suppresses
  21. L18:2210 weakly suppresses
  22. L16:1258 weakly excites
  23. L25:1135 weakly excites
  24. L19:2058 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 -0.73 and -0.671 values between -0.62 and -0.561 values between -0.56 and -0.512 values between -0.51 and -0.457 values between -0.45 and -0.3916 values between -0.39 and -0.3440 values between -0.34 and -0.2874 values between -0.28 and -0.23187 values between -0.23 and -0.17326 values between -0.17 and -0.12539 values between -0.12 and -0.06870 values between -0.06 and -0.00899 values between -0.00 and 0.05455 values between 0.05 and 0.11226 values between 0.11 and 0.16124 values between 0.16 and 0.2261 values between 0.22 and 0.2821 values between 0.28 and 0.3317 values between 0.33 and 0.399 values between 0.39 and 0.446 values between 0.44 and 0.506 values between 0.50 and 0.552 values between 0.55 and 0.614 values between 0.61 and 0.674 values between 0.67 and 0.723 values between 0.78 and 0.831 values between 0.83 and 0.891 values between 0.89 and 0.942 values between 0.94 and 1.001 values between 1.17 and 1.221 values between 1.22 and 1.281 values between 1.45 and 1.50offnegativepositiveoften

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. alties
  2. coming
  3. 时候
  4. Β
  5. ControlEvents
  6. getParameter
  7. aler
  8. 回事
  9. 些什么
  10. væ

Suppresses

  1. uo
  2. именно
  3. Product
  4. ioso
  5. Reflection
  6. }}
  7. Reflect
  8. 县政府
  9. Pope
  10. »