Neuron Explorer

Qwen3-0.6B

document end token

A neuron in layer 13 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 marker that separates one unrelated document from the next, especially after short lists or fragments of page text. It switches off on words cut in the middle of a phrase. Steering did not produce 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

    velocity (km/s)...?⏎- Inversely proportional⏎- Directly proportional⏎- Linearly related⏎- Non-linearly related⏎Similar Physics resources:<|endoftext|>RNA functions as the central conduit for information transfer in biology - in simple replicating

  2. #2 peak

    n");⏎- Compare two blocks of memory (function⏎- Locate first occurrence of character in string (function⏎- Locate last occurrence of character in string (function<|endoftext|>- Mattress Accessories⏎- Sleep Products⏎Also referred to as “responsible

  3. #3 peak

    , that's possible. 524 = 512.⏎The answer is 214 + 314 = 512<|endoftext|>[Haskell-beginners] Question about data structures⏎daniel.is.fischer at

  4. #4 peak

    , but pointers can be uninitialized, which is a nasty security hole. An eagerly evaluating functional language without hacks, would only allow for acyclic graphs of pointers.<|endoftext|>The Death And Life Of The Great Lakes Pdf⏎- and pdf⏎- Tuesday

  5. #5 peak

    ⏎July 08, 2009 |<|endoftext|>Computer Science Concentration for Students Matriculated After Jan. 1,

  6. #6 peak

    Mentor: Carl Ruetz⏎Page last modified July 14, 2009<|endoftext|>what does hamlet think about suicide? with textual support⏎2 Answers | Add

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 can I do to do it, can I do it. What can I do? What can I do? Maybe that's a problem. Let me try to think of different things. What can I do, maybe I can do that

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

    old stick to bash natural gas with is that nations will burn coal instead. Or just ignore the zealots and move forward rapidly to produce, export, and import methane. While Canada’s elite stands there slack-jawed going but what about our

  2. #2 peak

    child dies of a preventable cause, it is a huge tragedy and the nation has to get its act together and intensify efforts to make sure it doesn’t happen,” Anuradha Gupta, a senior official at the Health Ministry, said

  3. #3 peak

    percent of youths have witnessed cyberbullying and 49 percent of students have regretted something they have posted online.2 Girls more often resort to cyberbullying while guys still rule the physical bullyingworld on playgrounds and in school hallways

  4. #4 peak

    sorted and if not swapping randomly two elements. This is repeated until eventually the sequence is sorted.⏎If we consider all permutations of the first 4 natural numbers as input the expectation value of the number of swaps, averaged over all 4! input

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. L12:60 suppresses
  2. L8:1198 suppresses
  3. L12:262 weakly excites
  4. L12:1085 weakly excites
  5. L10:63 weakly excites
  6. L11:2128 weakly suppresses
  7. L11:1672 weakly excites
  8. L12:328 weakly suppresses
  9. L12:800 weakly excites
  10. L10:408 weakly excites
  11. L7:716 weakly excites
  12. L12:2052 weakly excites
  13. L9:2023 weakly suppresses
  14. L12:594 weakly excites
  15. L12:151 weakly suppresses
  16. L8:888 weakly excites
  17. L12:254 weakly suppresses
  18. L11:526 weakly excites
  19. L2:310 weakly suppresses
  20. L9:2275 weakly excites
  21. L11:1414 weakly excites
  22. L6:306 weakly excites
  23. L0:2805 weakly suppresses
  24. L9:40 weakly suppresses

Feeds later layers

  1. L14:1917 suppresses
  2. L14:179 excites
  3. L14:1660 suppresses
  4. L14:806 suppresses
  5. L14:1543 suppresses
  6. L14:256 suppresses
  7. L14:3011 suppresses
  8. L15:2488 excites
  9. L14:2133 suppresses
  10. L16:1425 suppresses
  11. L17:2879 suppresses
  12. L14:697 excites
  13. L14:2331 excites
  14. L14:1745 excites
  15. L15:2519 suppresses
  16. L15:2468 weakly excites
  17. L14:358 weakly suppresses
  18. L14:1902 weakly excites
  19. L15:2479 weakly suppresses
  20. L17:2499 weakly suppresses
  21. L14:2241 weakly suppresses
  22. L16:2497 weakly excites
  23. L14:2947 weakly suppresses
  24. L15:1147 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.

2 values between -1.80 and -1.681 values between -1.68 and -1.552 values between -1.16 and -1.032 values between -1.03 and -0.904 values between -0.90 and -0.774 values between -0.77 and -0.645 values between -0.64 and -0.5219 values between -0.52 and -0.3977 values between -0.39 and -0.26343 values between -0.26 and -0.131782 values between -0.13 and 0.001336 values between 0.00 and 0.13239 values between 0.13 and 0.2651 values between 0.26 and 0.3921 values between 0.39 and 0.524 values between 0.52 and 0.654 values between 0.65 and 0.773 values between 0.77 and 0.902 values between 0.90 and 1.031 values between 1.03 and 1.161 values between 1.16 and 1.291 values between 1.55 and 1.681 values between 1.68 and 1.811 values between 1.81 and 1.931 values between 1.93 and 2.061 values between 3.22 and 3.35offnegativepositiveoften

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. ##⏎⏎
  2. ',⏎⏎
  3. życie
  4. ’⏎⏎
  5. validated
  6. $⏎⏎
  7. ’.⏎⏎
  8. ';⏎⏎
  9. ';⏎⏎
  10. itle

Suppresses

  1. _periods
  2. 这方面
  3. 前三季度
  4. addObject
  5. ewire
  6. .getRight
  7. _REUSE
  8. enschaft
  9. 左右
  10. createForm