Neuron Explorer

Qwen3-0.6B

comma after qualifying adverbs

A neuron in layer 3 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 comma that comes right after a short phrase like "if so,", "or worse," or "more commonly," where the writer is about to add a qualifier or twist. When it was turned up, the text wandered off into unrelated rambling rather than showing a clear theme.

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

    church traditions where Lent is not observed at all. These folks may think of Lenten observance as, at best, a hollow custom, or, at worst, quite foreign to authentic Christianity. As a matter of fact, many who grew up

  2. #2 peak

    of conduct on those who dissent from them; to fetter the development, and, if possible, prevent the formation, of any individuality not in harmony with its ways, and

  3. #3 peak

    ate winter look!<|endoftext|>While doing my own comparison of women in the New Testament and Josephus’ Jewish histories, I began to notice a theme, or rather, two themes. Each one dealing with customs involving women and since I could not find

  4. #4 peak

    tolls than do wealthier people. But in the US, road systems are financed primarily through fuel taxes, vehicle registration fees, property taxes, and, increasingly, sales taxes—all of which are also regressive. Thus the relevant question

  5. #5 peak

    in Scotland originated with Anglo-Norman knights bearing the de Beauville or, more commonly, de Boyville name from Beauville, near Caen. They are believed to

  6. #6 peak

    any of these, you may have dry eye. A thorough ocular surface evaluation with our dry eye specialists will determine whether dry eye is present, and if so, what type(s) of dry eye are affecting

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 am" by Emily Brontstern? A. E. Brontstern A. E. Brontstern A. E. Brontstern A. E. Brontstern A

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

    . Men are 24 % less likely than women to visit their doctor for an annual checkup, even though their cancer mortality rates are far higher.⏎Moreover and as a matter of fact about men, let’s be honest. No man looks

  2. #2 peak

    fitness, it has also been shown to have other health benefits, including mental health benefits such as helping to manage depression and anxiety. Most importantly for many people, running is a good way to lose weight and generally get fitter

  3. #3 peak

    straight edge. the scrap ends left over from cutting a square from a rectangle work perfect. they can be used in two different ways, fractionally, and additively.⏎when working on any piece, it’s helpful to know where

  4. #4 peak

    une after bloom in late spring or summer. Wait until all chance of frost is past before pruning tropicals such as bougainvillea and hibiscus. Pruning stimulates tender new growth, which is susceptible to frost damage.⏎

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. L0:2633 weakly suppresses
  2. L2:86 weakly suppresses
  3. L1:198 weakly suppresses
  4. L2:2940 weakly excites
  5. L1:2336 weakly excites
  6. L2:1809 weakly suppresses
  7. L0:2526 weakly suppresses
  8. L0:2495 weakly suppresses
  9. L0:224 weakly suppresses
  10. L1:2353 weakly suppresses
  11. L1:2417 weakly suppresses
  12. L1:465 weakly excites
  13. L0:1873 weakly excites
  14. L1:1426 weakly suppresses
  15. L0:3013 weakly suppresses
  16. L2:3062 weakly excites
  17. L2:2425 weakly excites
  18. L0:1342 weakly excites
  19. L2:2915 weakly excites
  20. L2:670 weakly excites
  21. L1:1889 weakly suppresses
  22. L2:1036 weakly excites
  23. L1:2625 weakly excites
  24. L0:1189 weakly suppresses

Feeds later layers

  1. L4:1444 excites
  2. L6:918 weakly excites
  3. L8:611 weakly excites
  4. L15:2835 weakly suppresses
  5. L9:2771 weakly suppresses
  6. L5:540 weakly excites
  7. L6:2575 weakly excites
  8. L4:1441 weakly excites
  9. L4:2610 weakly suppresses
  10. L12:2332 weakly suppresses
  11. L17:794 weakly suppresses
  12. L13:1706 weakly suppresses
  13. L20:2065 weakly suppresses
  14. L17:2891 weakly excites
  15. L16:1993 weakly suppresses
  16. L15:396 weakly excites
  17. L6:2147 weakly suppresses
  18. L9:2769 weakly suppresses
  19. L8:852 weakly suppresses comma inside large numbers
  20. L21:1643 weakly excites
  21. L4:610 weakly suppresses
  22. L6:137 weakly suppresses
  23. L4:744 weakly suppresses
  24. L4:810 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 -2.03 and -1.831 values between -1.63 and -1.421 values between -1.42 and -1.221 values between -1.22 and -1.024 values between -1.02 and -0.8136 values between -0.81 and -0.6187 values between -0.61 and -0.41147 values between -0.41 and -0.211512 values between -0.21 and -0.001944 values between -0.00 and 0.20132 values between 0.20 and 0.4012 values between 0.40 and 0.616 values between 0.61 and 0.815 values between 1.01 and 1.212 values between 1.21 and 1.421 values between 1.42 and 1.621 values between 1.62 and 1.823 values between 1.82 and 2.032 values between 2.03 and 2.231 values between 2.23 and 2.433 values between 2.43 and 2.631 values between 2.63 and 2.841 values between 3.65 and 3.851 values between 3.85 and 4.052 values between 5.27 and 5.471 values between 5.88 and 6.08offnegativepositiveoften

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. essenger
  2. semblies
  3. plementation
  4. ístico
  5. ово
  6. úmero
  7. asta
  8. PLICATION
  9. structor
  10. plement

Suppresses

  1. etine
  2. 蓐
  3. ()`
  4. /$',
  5. ()'
  6. %'
  7. ()%
  8. ?>"><?
  9. `⏎⏎
  10. apanese