Neuron Explorer

Qwen3-0.6B

unclear: electrons and code references

A neuron in layer 7 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 a mixed set of words, mostly 'electron' and 'pattern matching', plus 'reference' and 'local variable'. It is not obvious what links them. Turning it up gave answers that wandered without a clear theme.

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

    temperatures are the factors of electrophilic substitution on aromatic compounds.⏎The most common reaction of aromatic compounds is electrophilic substitution. It is when an electron-poor reagent (an electrophile, E+) reacts with the electron

  2. #2 peak

    , respectively), but⏎Bool doesn’t have such a function. All⏎Bool is stuck with is the build-in⏎if then else-syntax, or worse: pattern matching.⏎With a ‘fold’ function (also known as

  3. #3 peak

    . It is the number of protons in the nucleus minus the number of electrons in between the nucleus and the electron in question. Basically, the nucleus attracts an electron, but other electrons in lower shells repel it (opposites attract, likes

  4. #4 peak

    shielding effect is. Essentially, the core electrons shield the valence electrons from the positive charge of the nucleus.⏎Electron-Electron Repulsions ⏎When two electrons are in the same shell, they will repel each other slightly. This effect

  5. #5 peak

    believe that pattern matching has been proved to be a useful and expressive tool in various modern languages. In particular, many aspects of this PEP were inspired by how pattern matching works in Rust and Scala .⏎Both those languages are statically typed, which

  6. #6 peak

    setting. This repair method⏎The above information equips you with everything you need to know about FCAW as a new welder. The techniques and advantages and disadvantages of FCA

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: "The problem of the elderly" (Problem of the Elderly) with regard to the elderly and the management staff. The Problem of the Elderly is a common problem. Please, the management staff, which is for the people. I

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

    residence.⏎Having a professionally installed, monitored fire alarm system in your home offers you the best overall protection. When a fire is sensed, your alarm will activate all evacuation horns throughout your home. Even if you are on another level of your home,

  2. #2 peak

    , provide a source of natural ventilation, and can become a lifesaver in an emergency. If there is a fire in your home, for example, a single stairwell may become unusable, leaving family members and tenants without a clear path to

  3. #3 peak

    - The boy, that was playing in the street, that got hit by a car was my brother.⏎- My neighbor, that lives on the 2nd floor, is going to take me to school tomorrow.<|endoftext|>In the early 1

  4. #4 peak

    won’t bite you, but this still doesn’t reduce your anxiety. Likewise, you may know that it is safe to be out on a balcony in a high-rise block, yet, feeling terrified to go out on it or even enjoy the view

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. L4:1812 suppresses
  2. L6:1144 excites
  3. L6:681 excites
  4. L6:2557 weakly suppresses
  5. L2:1657 weakly excites
  6. L3:1239 weakly suppresses
  7. L1:2948 weakly excites
  8. L6:776 weakly suppresses
  9. L6:983 weakly excites
  10. L4:2708 weakly suppresses
  11. L6:1432 weakly excites
  12. L6:1659 weakly excites
  13. L6:953 weakly suppresses
  14. L6:531 weakly suppresses
  15. L6:2868 weakly excites
  16. L6:2584 weakly suppresses
  17. L2:366 weakly suppresses
  18. L6:744 weakly excites
  19. L4:903 weakly suppresses
  20. L6:2110 weakly suppresses
  21. L6:1733 weakly suppresses
  22. L6:1355 weakly excites
  23. L5:1950 weakly suppresses unclear: 'Lorenz' and name fragments
  24. L6:1420 weakly suppresses

Feeds later layers

  1. L10:1563 suppresses
  2. L20:690 excites
  3. L8:2762 weakly excites
  4. L10:2572 weakly excites
  5. L21:2226 weakly excites
  6. L22:2553 weakly excites
  7. L25:1190 weakly excites
  8. L17:687 weakly excites
  9. L24:3049 weakly excites
  10. L9:1814 weakly suppresses
  11. L9:711 weakly excites
  12. L9:2832 weakly suppresses
  13. L25:2315 weakly suppresses
  14. L25:1065 weakly suppresses
  15. L10:1864 weakly excites
  16. L8:1687 weakly suppresses
  17. L21:1891 weakly excites
  18. L8:2549 weakly excites
  19. L26:746 weakly suppresses unclear: mixed word-start fragments
  20. L25:975 weakly excites
  21. L15:1133 weakly excites
  22. L20:876 weakly suppresses
  23. L17:2837 weakly excites
  24. L9:1392 weakly 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.

1 values between -5.59 and -5.391 values between -2.07 and -1.871 values between -1.09 and -0.892 values between -0.89 and -0.705 values between -0.70 and -0.5068 values between -0.50 and -0.31735 values between -0.31 and -0.112215 values between -0.11 and 0.08781 values between 0.08 and 0.2892 values between 0.28 and 0.482 values between 0.48 and 0.672 values between 0.87 and 1.061 values between 1.06 and 1.261 values between 1.65 and 1.841 values between 2.04 and 2.23offnegativepositiveoften

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. рон
  2. 去了
  3. -binding
  4. ythe
  5. ariat
  6. zar
  7. ��
  8. zek
  9. inston
  10. aleza

Suppresses

  1. 读后
  2. ulates
  3. ,),⏎
  4. walls
  5. slime
  6. .getContent
  7. asters
  8. 涞
  9. duct
  10. 一期