Neuron Explorer

Qwen3-0.6B

'and' joining a pair of options

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 lights up on the word 'and' (or 'or') that links two halves of a pair, especially after phrases like 'two key factors' or 'two models'. Turning it up made the model produce loops of text, so the steering adds little here.

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

    it in the START menu on the bottom left corner of your computer screen.Ad⏎2Go to modelspace. There are two views in AutoCAD: modelspace and paperspace. Your drawing should always be made in modelspace, and dimensions that

  2. #2 peak

    the substrate that it will catalyse. There are two models that describe the relationship between the active site and the substrate and they are the lock and key model and induced fit hypothesis.⏎- The shape of the active site is determined primarily by the

  3. #3 peak

    pair.<|endoftext|>This is a blog post on how to hook up or connect your battery. This article will be broken into two parts, the "How-to" and the "Why." The How-to section features detailed instructions with illustrations, including what

  4. #4 peak

    women and girls.<|endoftext|>Microcontroller vs FPGA: Understanding the Key Differences⏎When it comes to embedded systems design, two technologies that often come up are microcontrollers and field-programmable gate arrays (FPGAs). While both serve similar purposes

  5. #5 peak

    ’s job.⏎See how to implement NIST CSF<|endoftext|>Minimum & single dose⏎In homeopathy two important principles are employed when prescribing medicines: the minimum dose and the single dose.⏎The principle that only ‘the minimum dose’ should be employed

  6. #6 peak

    spindle?⏎At present, the primary control of the FANUC system’s spindle motor has two interfaces: analog (DC0 ~ 10v) interface and digital (serial transmission) interface. The analog interface is driven by a frequency converter

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: if I have a question, I will say that question, and I will say that question, and I will say that question. and this question, and I will say that question, and this question, and I will say that question,

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

    of Minnesota students. You can see all of the details involved in making it in the video below.<|endoftext|>Create a Graph⏎Library Home || Full Table of Contents || Suggest a Link || Library Help⏎|National Center for Education Statistics (NC

  2. #2 peak

    the raw materials used in production. Such<|endoftext|>The Leading eBooks Store Online⏎for Kindle Fire, Apple, Android, Nook, Kobo, PC, Mac, BlackBerry ...⏎DK Eyewitness Books: Gorilla⏎Here is a spectacular and informative

  3. #3 peak

    a matrix⏎Find nth maximum⏎Matlab Basics II - Create a vector with a repeated entry⏎Create sequnce 1 4 9 16 25.........⏎Basics: Divide integers to get

  4. #4 peak

    ocean...⏎and hydrologic sciences⏎Hydrology is the study of the movement, distribution, and quality of water on Earth and other planets, including the hydrologic cycle, water

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. L12:51 excites
  2. L12:2808 suppresses
  3. L12:1235 suppresses commas inside numbers and lists
  4. L12:2548 excites
  5. L12:1212 suppresses
  6. L12:1197 excites
  7. L9:212 excites
  8. L12:1095 suppresses
  9. L8:131 suppresses
  10. L8:70 excites 'and' between stacked superlatives
  11. L12:1246 suppresses
  12. L9:1449 suppresses
  13. L8:809 excites
  14. L12:2409 excites
  15. L12:2934 suppresses
  16. L12:2171 excites
  17. L11:1431 excites commas and 'and' in noun lists
  18. L12:1062 suppresses
  19. L12:2058 excites
  20. L6:1084 suppresses
  21. L12:1629 excites
  22. L9:1820 excites
  23. L12:2891 excites
  24. L8:185 suppresses

Feeds later layers

  1. L15:1948 excites
  2. L14:783 excites
  3. L16:1074 excites
  4. L14:672 excites
  5. L15:3062 excites
  6. L16:925 excites
  7. L15:611 excites
  8. L14:1060 suppresses
  9. L16:316 excites
  10. L14:491 suppresses
  11. L15:989 weakly suppresses
  12. L14:686 weakly suppresses
  13. L18:1013 weakly excites
  14. L14:747 weakly suppresses
  15. L16:1294 weakly excites
  16. L14:53 weakly excites
  17. L16:2177 weakly suppresses
  18. L14:453 weakly suppresses
  19. L17:747 weakly suppresses
  20. L15:2476 weakly suppresses
  21. L16:340 weakly excites
  22. L14:660 weakly suppresses
  23. L15:2449 weakly suppresses line breaks mid-sentence
  24. L22:637 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 -1.42 and -1.231 values between -1.23 and -1.032 values between -1.03 and -0.849 values between -0.84 and -0.64106 values between -0.64 and -0.45550 values between -0.45 and -0.251543 values between -0.25 and -0.051330 values between -0.05 and 0.14143 values between 0.14 and 0.3448 values between 0.34 and 0.5322 values between 0.53 and 0.7319 values between 0.73 and 0.9216 values between 0.92 and 1.1215 values between 1.12 and 1.3219 values between 1.32 and 1.5113 values between 1.51 and 1.7112 values between 1.71 and 1.906 values between 1.90 and 2.108 values between 2.10 and 2.294 values between 2.29 and 2.499 values between 2.49 and 2.686 values between 2.68 and 2.887 values between 2.88 and 3.072 values between 3.07 and 3.275 values between 3.27 and 3.471 values between 3.47 and 3.661 values between 3.66 and 3.864 values between 3.86 and 4.051 values between 4.25 and 4.442 values between 4.44 and 4.641 values between 4.83 and 5.031 values between 5.81 and 6.011 values between 6.20 and 6.40offnegativepositiveoften

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. erval
  2. TargetException
  3. verständlich
  4. uracy
  5. 个工作
  6. eder
  7. ometric
  8. ierte
  9. edBy
  10. ted

Suppresses

  1. yll
  2. 纲
  3. ласт
  4. 法则
  5. đích
  6. олос
  7. -class
  8. ฎ
  9. ells
  10. rün