Neuron Explorer

Qwen3-0.6B

full stop ending boilerplate or a title

A neuron in layer 10 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 reacts to the full stop that closes a short line, such as the dot in 'This preview shows pages 1 to 2.' or the end of a title. When it was turned up, the model's text did not show 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

    is not a viable option, as it is a highly-energy intensive process, and half or more of the water produced<|endoftext|>This preview shows pages 1–2. Sign up to view the full content.⏎

  2. #2 peak

    caused<|endoftext|>This preview shows pages 1–2. Sign up to view the full content.⏎This preview has intentionally blurred sections. Sign

  3. #3 peak

    conservative views that are particularly detrimental for women.⏎But nothing — not even a conservative culture or fear of the Taliban<|endoftext|>This preview shows pages 1–2. Sign up to view the full content.⏎This preview has intentionally blurred sections. Sign

  4. #4 peak

    SciPy or R programs.<|endoftext|>by Roman Davidov⏎(revisited from the June 2003 issue of Meteorite Times)⏎Zhamanshin. This word translated from Kazakh language like “bad soil” or “land without

  5. #5 peak

    Research Laboratory, Pittsburgh, PA 15236⏎Other Occupational Concerns⏎Longwall USA, International Exhibition & Conference<|endoftext|>Hypertext Markup Language. The standard language used to create Web pages. Originally based on SGML, note

  6. #6 peak

    in synthesis Home⏎PDF. Rich HTML. We have demonstrated purely mechanochemical hydrogenation of MgB2

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: about in 1995-2000 in the first year of work at the 2000 research in the 1900s in the 1990s in the 190

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

    scent of baby shampoo under your nose? Most of us think such moments are gone after our snuggler learns to read. But they don’t have to be! Save the ritual, and you can help her become an even better reader than she

  2. #2 peak

    ery, and power politics—all wrapped up in a story befitting a good dime store novel. And the proverbial plot thickened when computer scientists got involved—at that point, things got downright ugly. As painful as it was for the computer

  3. #3 peak

    over?⏎Quite simply, if you actually look at the broad measures of aggregate economic activity that define recessions and recoveries, they don’t look like those letters—at best they would be Procrustean fits. For example, the Great Depression

  4. #4 peak

    because dopamine connects neurons, which helps you recognize everything associated with a rewarding experience. The tribe is thrilled by the fish feast—for a few days. But then the dopamine sags. So you go out for

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. L6:250 excites
  2. L4:7 excites
  3. L9:819 suppresses
  4. L4:3 excites commas after citation fields
  5. L9:458 excites
  6. L5:168 suppresses
  7. L4:1546 suppresses
  8. L5:974 weakly excites
  9. L9:1802 weakly excites
  10. L8:517 weakly suppresses
  11. L5:1653 weakly excites
  12. L3:487 weakly suppresses
  13. L9:1113 weakly suppresses
  14. L9:16 weakly suppresses unclear: periods inside numbers and names
  15. L7:948 weakly suppresses
  16. L5:454 weakly suppresses
  17. L6:2621 weakly suppresses
  18. L3:640 weakly suppresses
  19. L8:93 weakly suppresses period after a middle initial
  20. L1:2157 weakly excites
  21. L8:2184 weakly excites
  22. L8:573 weakly excites
  23. L8:1578 weakly suppresses
  24. L9:121 weakly suppresses

Feeds later layers

  1. L16:1112 excites
  2. L14:155 excites sentence-ending periods in prose
  3. L17:1772 excites
  4. L17:1105 excites
  5. L21:1086 excites end of sentence before a claim or statistic
  6. L14:153 excites
  7. L13:9 suppresses
  8. L17:990 excites
  9. L16:241 suppresses
  10. L11:1272 excites unclear: sentence ends in factual prose
  11. L24:148 excites
  12. L14:409 suppresses
  13. L16:436 excites
  14. L13:345 excites
  15. L17:1770 excites
  16. L11:132 excites period ending an opening topic sentence
  17. L19:2097 excites sentence-end punctuation in speech
  18. L12:771 excites
  19. L16:2535 excites
  20. L17:839 excites
  21. L17:2613 excites
  22. L19:2614 excites
  23. L16:1545 excites
  24. L16:1082 excites comma linking two clauses after 'and'

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 -1.96 and -1.851 values between -1.85 and -1.731 values between -1.73 and -1.621 values between -1.62 and -1.502 values between -1.16 and -1.051 values between -1.05 and -0.932 values between -0.93 and -0.822 values between -0.82 and -0.709 values between -0.70 and -0.5928 values between -0.59 and -0.4876 values between -0.48 and -0.36275 values between -0.36 and -0.25760 values between -0.25 and -0.131282 values between -0.13 and -0.021067 values between -0.02 and 0.10220 values between 0.10 and 0.2166 values between 0.21 and 0.3328 values between 0.33 and 0.4421 values between 0.44 and 0.5515 values between 0.55 and 0.678 values between 0.67 and 0.788 values between 0.78 and 0.904 values between 0.90 and 1.017 values between 1.01 and 1.133 values between 1.13 and 1.245 values between 1.24 and 1.364 values between 1.36 and 1.472 values between 1.47 and 1.582 values between 1.58 and 1.701 values between 1.70 and 1.812 values between 1.93 and 2.041 values between 2.39 and 2.503 values between 2.50 and 2.61offnegativepositiveoften

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. elay
  2. roker
  3. whereas
  4. enez
  5. кроме
  6. bra
  7. 而在
  8. ollo
  9. submit
  10. 有条件

Suppresses

  1. t
  2. political
  3. ander
  4. orestation
  5. 岚
  6. oat
  7. Meal
  8. /L
  9. is
  10. -te