Neuron Explorer

Qwen3-0.6B

unclear: mid-word pieces inside names

A neuron in layer 6 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 is weak and hard to read. Its highest reactions fall on pieces of longer names or technical words, like the middle of a species name or a place name split into parts. Turning it up or down didn't produce 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

    of the family Veneridae Rafinesque, 1815. Three species, including one new, are here described: Goshoraia minor Tashiro and Kozai, 1989 (Apt

  2. #2 peak

    in this way."⏎In their paper, published in the current issue of the Public Library of Science ONE, they say there's an established link between Type 1 diabetes and the body's ability to

  3. #3 peak

    ship shape = girlfriend friendship shipshape.⏎waist ? tail ? ? side ? ? fall ? ? down ? ? spring ? ? ? hole⏎Answerwaist coat tail gate way side arm pit fall out put down cast off spring time line man

  4. #4 peak

    referred to as Wolbachia is Wolbachia pipientis. Before DNA sequencing, when bacteria were classified morphologically, W. persica and W. melaphagi were also thought to be related, but these have nothing to do with

  5. #5 peak

    as:⏎- Rickets caused by vitamin D deficiency⏎- Pellagra caused by niacin deficiency⏎- Scurvy caused by vitamin C deficiency⏎- Beriberi caused by vitamin B1 (thiamine) deficiency⏎In the advent

  6. #6 peak

    . Latvia's neighbours to the east are the countries Russia and Belarus. Latvia is split into four parts called Kurzeme, Vidzeme, Zemgale, and Latgale.⏎History[change | change source]⏎Latvia was

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 have been thinking about lately, I have been thinking about lately, I have been thinking about lately, I have been thinking about lately, I have been thinking about lately, I have been thinking about lately, I have been thinking about lately

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

    uses the resulting Internet address to connect to the web server, send the e-mail or use the World Wide Web.⏎When you visit http://www.internetsociety.org in a browser,

  2. #2 peak

    in space in detail, but they do indirectly contemplate the application of the law of armed conflict. Both the constitutive legal instruments and the subsequent efforts to address the current challenges to space security influence the way in which the law of

  3. #3 peak

    the process of amputating or cutting off a portion or all of the last bone of each digit/toe of the paw while the cat is under anesthesia.1-3 This is usually done at the pet owner’s request to prevent cats from

  4. #4 peak

    SYNDROME X DO? What does Metabolic Syndrome X cause? What are the effects of Metabolic Syndrome X? Over time, normally years, the metabolic imbalances give rise to a greater risk of cardiovascular disease, heart disease, kidney failure

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. L4:278 strongly suppresses
  2. L2:39 strongly excites
  3. L2:69 strongly excites
  4. L4:129 strongly suppresses
  5. L1:140 strongly suppresses
  6. L4:695 excites
  7. L3:54 excites
  8. L3:280 excites
  9. L2:60 excites
  10. L0:144 suppresses
  11. L5:1341 suppresses
  12. L1:981 suppresses
  13. L4:436 excites
  14. L2:143 suppresses
  15. L2:33 excites
  16. L1:647 suppresses
  17. L1:434 suppresses
  18. L1:80 suppresses
  19. L2:195 excites
  20. L3:593 excites
  21. L5:29 suppresses
  22. L1:698 suppresses
  23. L0:96 suppresses
  24. L0:309 suppresses

Feeds later layers

  1. L9:1574 strongly excites
  2. L11:1243 strongly excites
  3. L7:461 strongly excites
  4. L12:831 strongly excites
  5. L10:1509 strongly excites
  6. L14:74 strongly excites
  7. L15:1550 strongly excites
  8. L8:154 strongly suppresses
  9. L9:421 strongly suppresses
  10. L12:1217 strongly excites
  11. L16:841 suppresses
  12. L8:654 excites
  13. L15:1583 suppresses
  14. L18:688 excites
  15. L12:396 suppresses
  16. L15:1058 suppresses
  17. L16:409 excites
  18. L8:630 suppresses
  19. L14:210 excites
  20. L16:654 excites
  21. L12:619 suppresses
  22. L16:766 suppresses
  23. L8:124 excites
  24. L9:342 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 -2.57 and -2.491 values between -2.01 and -1.931 values between -1.77 and -1.691 values between -1.69 and -1.611 values between -1.61 and -1.531 values between -1.21 and -1.132 values between -1.13 and -1.054 values between -1.05 and -0.976 values between -0.97 and -0.893 values between -0.89 and -0.818 values between -0.81 and -0.735 values between -0.73 and -0.658 values between -0.65 and -0.578 values between -0.57 and -0.4915 values between -0.49 and -0.4122 values between -0.41 and -0.3360 values between -0.33 and -0.25111 values between -0.25 and -0.17373 values between -0.17 and -0.091156 values between -0.09 and -0.011685 values between -0.01 and 0.07304 values between 0.07 and 0.1584 values between 0.15 and 0.2334 values between 0.23 and 0.3110 values between 0.31 and 0.392 values between 0.39 and 0.472 values between 0.55 and 0.63offnegativepositiveoften

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. ros
  2. ROT
  3. syn
  4. _SOL
  5. ROL
  6. sol
  7. rays
  8. uada
  9. ALAR
  10. AMPL

Suppresses

  1. iki
  2. .dispose
  3. iver
  4. jak
  5. .escape
  6. hiểm
  7. aşağıd
  8. cube
  9. adjud
  10. —we