Neuron Explorer

Qwen3-0.6B

"how" after learn or know

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 on the word "how" when it follows verbs like "learn", "know" or "determine", as in "learn how to play". Turning it up did not clearly bring out that theme in the sample writing, though the examples are very consistent.

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

    tell that an essay has been plagiarized, even if unsure of the exact source of pirated passages. Plagiarism software can be used to help determine how original - or unoriginal - a student's written work may be. A student

  2. #2 peak

    listening to your favorite pieces -- a real extra dimension for your music enjoyment.⏎Later you will also be able to use your ability to read music notes if you learn how to play a musical instrument.⏎How to Read Music Notes - Introduction⏎No comments

  3. #3 peak

    them to new textures and sights during their tummy time.⏎You can place their favorite toy in front of them so that they can practice their motor skills and learn how to grab.⏎Tummy time fosters independence and a sense of learning in a

  4. #4 peak

    grade children should be giving consideration and paying more attention to the correct spelling of words. They should be spelling previously learned words correctly. If the child does not know how to spell a word, he or she

  5. #5 peak

    of the misconceptions spread about it by its critics - Angst… that the Energiewende will work.⏎Unlike the Swedes and Italians, the Germans knew how to replace nuclear back in 2002 – and they have even replaced

  6. #6 peak

    and adopted each year and will penalize laboratories that do not comply with adoption standards. Currently, labs are required to report to the U.S. Department of Agriculture how many animals are used each year, but these reports can be misleading. USDA reports

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 have been able to take some ideas and say a more concrete and more conscious awareness awareness awareness awareness awareness awareness awareness awareness awareness awareness awareness awareness awareness awareness awareness awareness awareness awareness awareness awareness awareness awareness awareness awareness awareness awareness awareness awareness awareness awareness awareness awareness

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

    ’s super easy to get started too.⏎Here’s a quick how-to:⏎- Get in a comfy position. You don’t need to pretzel it up, know lying down can lead to sleeping, but other than that, just be comfortable in

  2. #2 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,

  3. #3 peak

    I’m Jane Houston Jones at NASA’s Jet Propulsion Laboratory in Pasadena, California.⏎Have you ever seen the Space Station pass overhead? It’s easy if you know when and where to look.⏎The Space Station orbits Earth about 15 times

  4. #4 peak

    said Guillaume Gronoff, a French astrophysicist at NASA Langley.⏎"And you think, 'OK, what I am looking at — it's a<|endoftext|>An ultra-Orth

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. L16:2097 suppresses
  2. L16:1556 suppresses
  3. L17:913 excites
  4. L15:2392 suppresses
  5. L14:2142 suppresses
  6. L17:2582 excites
  7. L11:1469 excites
  8. L10:1919 suppresses
  9. L17:1952 excites
  10. L11:1774 excites
  11. L14:1720 excites
  12. L17:2772 suppresses
  13. L14:462 excites
  14. L17:1562 suppresses
  15. L15:2053 excites
  16. L16:2718 suppresses
  17. L16:1672 excites
  18. L15:433 excites
  19. L14:2460 weakly suppresses
  20. L13:1311 weakly suppresses
  21. L13:1776 weakly excites
  22. L10:254 weakly suppresses
  23. L17:1321 weakly suppresses
  24. L9:1453 weakly excites

Feeds later layers

  1. L19:1631 suppresses
  2. L19:1808 excites
  3. L21:918 excites
  4. L19:1483 suppresses
  5. L20:2495 suppresses
  6. L19:1548 excites
  7. L19:1177 excites
  8. L19:1824 suppresses
  9. L20:858 suppresses
  10. L19:1796 excites
  11. L19:2143 suppresses
  12. L22:3009 suppresses
  13. L19:580 excites
  14. L23:594 excites
  15. L22:2067 suppresses
  16. L19:2643 weakly excites "is" in "what is" questions
  17. L20:330 weakly excites
  18. L19:1055 weakly excites
  19. L22:543 weakly excites
  20. L19:1396 weakly suppresses
  21. L21:1910 weakly suppresses
  22. L20:2433 weakly suppresses
  23. L19:2117 weakly suppresses
  24. L20:1413 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 -4.79 and -4.301 values between -4.30 and -3.812 values between -3.81 and -3.333 values between -2.84 and -2.358 values between -2.35 and -1.8721 values between -1.87 and -1.3881 values between -1.38 and -0.89342 values between -0.89 and -0.412338 values between -0.41 and 0.08790 values between 0.08 and 0.57108 values between 0.57 and 1.0562 values between 1.05 and 1.5441 values between 1.54 and 2.0325 values between 2.03 and 2.5123 values between 2.51 and 3.009 values between 3.00 and 3.4911 values between 3.49 and 3.978 values between 3.97 and 4.467 values between 4.46 and 4.959 values between 4.95 and 5.435 values between 5.43 and 5.922 values between 5.92 and 6.414 values between 6.41 and 6.892 values between 6.89 and 7.382 values between 7.87 and 8.351 values between 8.84 and 9.331 values between 9.81 and 10.301 values between 14.19 and 14.68offnegativepositiveoften

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. 为何
  2. why
  3. ela
  4. iele
  5. why
  6. how
  7. 何时
  8. 為什麼
  9. يا
  10. 如何

Suppresses

  1. 的认识
  2. 知道
  3. 的理解
  4. 描述
  5. 了解到
  6. 了解
  7. 启示
  8. identified
  9. asurer
  10. 理解和