Neuron Explorer

Qwen3-0.6B

abstract head nouns after 'the' or 'some'

A neuron in layer 12 of Qwen3-0.6B. Its strongest snippets are mixed: they peak on different tokens. It is featured, so it has steering outputs.

What this neuron does

This neuron fires on abstract nouns that complete a phrase, such as 'law' in 'supreme law', 'faith' in 'Catholic faith', or 'knowledge' after 'some'. It is most negative on words like 'themselves' and 'one another'. Turning it up did not give a clear, matching change in the text.

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

    it is 180°, and for a class C amplifier it is less than 180°.<|endoftext|>The Constitution of Nigeria is the supreme law of Nigeria. It sets out the framework for government and all its activities, including

  2. #2 peak

    essential qualities and characteristics which are transmitted over time.I am reminded of GK Chesterton's notion of The Thing. The idea that there is some thing called the Catholic faith and it makes sense of every other thing. There is not one area of life

  3. #3 peak

    Gertrude has some knowledge of what happened, whether she is mentally unbalanced and is in denial, or

  4. #4 peak

    improve your students' grades. In the Course Activity Related to Grades report, also known as learning analytics, you can view how active your students are in your course materials. This activity has been shown to correlate strongly with course outcomes. Students who are

  5. #5 peak

    most prominent unique personal characteristics of the G&T are: being Critical, Persistent & Independent of thought & judgement. Studies conclude that there are at least 18 areas of sensitivity. (CHART by Dunn & Price)⏎Also see posts: “

  6. #6 peak

    's Medicare and Medicaid programs are willing to cover. Lichtenberg writes that there are two ways to improve health-care quality: by ensuring that patients get "best practice" care and by "shifting the technological frontier" through the use of new

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 think the problem with the problem of the problem of the problem of the problem of the problem. The problem with the problem. The problem of the problem. The problem with the problem. The problem of the problem. The problem.

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

    the church? This question, frequently discussed in THE EXAMINER, is important because it addresses the basic issues of Christlike relationships, Christlike respect for one another, Christlike service of the powerful to the powerless, and the individual responsibility

  2. #2 peak

    was born.⏎Cloning is not new to nature, just new to humanity. Many creatures practice asexual reproduction, or parthenogenesis, producing exact copies of themselves. Many plants reproduce themselves by sending out roots

  3. #3 peak

    in thinking it made the speaker sound self-righteous.⏎Mark, here are your answers.⏎Illeism Is the Habit of Referring to Yourself in the Third Person⏎This verbal tic is known as "illeism." That's the habit of

  4. #4 peak

    we want to retrieve the correct value from the data table above.⏎Have you ever built an app based on Excel?⏎Method 1: The easiest way (VLOOKUP & MATCH)⏎Probably the easiest – or at least the shortest – way of

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. L10:850 suppresses
  2. L10:2679 suppresses
  3. L11:305 suppresses
  4. L10:164 excites
  5. L10:483 excites
  6. L11:1482 excites
  7. L11:2545 excites
  8. L9:64 excites
  9. L2:558 suppresses
  10. L11:420 excites
  11. L10:1390 weakly suppresses
  12. L10:324 weakly suppresses
  13. L10:2645 weakly excites
  14. L7:1402 weakly excites
  15. L9:868 weakly suppresses
  16. L10:1103 weakly suppresses
  17. L11:865 weakly excites
  18. L2:314 weakly suppresses
  19. L9:383 weakly suppresses
  20. L11:1542 weakly suppresses
  21. L10:2675 weakly excites
  22. L10:2431 weakly excites
  23. L6:2652 weakly excites
  24. L9:63 weakly excites

Feeds later layers

  1. L15:837 strongly excites
  2. L13:1933 suppresses
  3. L14:1651 excites
  4. L14:1364 suppresses
  5. L13:2783 excites
  6. L14:2126 excites
  7. L13:1667 excites
  8. L13:1807 suppresses
  9. L14:525 suppresses
  10. L16:335 suppresses
  11. L13:971 excites
  12. L13:1470 excites
  13. L14:1109 excites
  14. L16:1144 excites
  15. L13:2133 excites
  16. L13:167 suppresses
  17. L15:878 suppresses
  18. L13:1843 excites
  19. L13:193 suppresses
  20. L13:1073 excites
  21. L14:868 excites
  22. L13:2056 suppresses
  23. L14:576 excites
  24. L21:2746 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 -2.53 and -2.421 values between -2.21 and -2.101 values between -2.10 and -2.001 values between -2.00 and -1.892 values between -1.89 and -1.781 values between -1.68 and -1.574 values between -1.57 and -1.464 values between -1.46 and -1.363 values between -1.36 and -1.252 values between -1.25 and -1.154 values between -1.15 and -1.0411 values between -1.04 and -0.934 values between -0.93 and -0.8311 values between -0.83 and -0.729 values between -0.72 and -0.6128 values between -0.61 and -0.5145 values between -0.51 and -0.4095 values between -0.40 and -0.29240 values between -0.29 and -0.19566 values between -0.19 and -0.081539 values between -0.08 and 0.03700 values between 0.03 and 0.13264 values between 0.13 and 0.24140 values between 0.24 and 0.3582 values between 0.35 and 0.4538 values between 0.45 and 0.5639 values between 0.56 and 0.6726 values between 0.67 and 0.7715 values between 0.77 and 0.8810 values between 0.88 and 0.996 values between 0.99 and 1.093 values between 1.09 and 1.202 values between 1.20 and 1.304 values between 1.30 and 1.411 values between 1.41 and 1.523 values between 1.52 and 1.622 values between 1.62 and 1.73offnegativepositiveoften

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. trì
  2. newsp
  3. jective
  4. idata
  5. 面前
  6. Quốc
  7. mers
  8. cı
  9. »⏎⏎
  10. 这支

Suppresses

  1. 矣
  2. uds
  3. いません
  4. ubishi
  5. Raised
  6. ),$
  7. auled
  8. raised
  9. Raised
  10. elt