Neuron Explorer

Qwen3-0.6B

joins inside names and decimal digits

A neuron in layer 27 of Qwen3-0.6B. Its strongest snippets are fairly consistent: several of them peak on the same token. It is featured, so it has steering outputs.

What this neuron does

This neuron fires at points inside a name or term where a hyphen, bracket or 'of' links two parts, like the bracket in 'Wired Equivalent Privacy (WEP)'. It also goes strongly negative on digits inside decimal numbers. The steered outputs were mixed and did not show one 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

    developed by the Wi-Fi Alliance to secure wireless computer networks. The Alliance defined these in response to serious weaknesses researchers had found in the previous system, Wired Equivalent Privacy (WEP).⏎WPA (sometimes referred to as the draft IEEE

  2. #2 peak

    learn how to use different art mediums and offer your child the opportunity for self-expression. Activities include scribbling with crayons, chalking, clay or Play-doh, easel painting, and watercolor painting.⏎Side Note: Gluing

  3. #3 peak

    series so PLEASE, PLEASE leave a comment if you LOVE your math series! I'd love to check it out!⏎I like the tricks (9s trick, Doubles, Doubles Plus 1) that Saxon uses to teach facts,

  4. #4 peak

    a rhyme⏎to rock a rhyme that’s right on time⏎It’s tricky… it’s tricky (tricky) tricky (tricky)”⏎–Run-D.M.C., “It’s Tricky”⏎Long-term planning is hard, especially

  5. #5 peak

    AYCAT.⏎As part of The World As It Could Be Tree Dedication Ceremony, students, teachers, and other collaborators on the project reflect on the Universal Declaration of Human Rights. Balboa High School Tree Ceremony from BAYCAT.<|endoftext|>Several

  6. #6 peak

    have been missing. It helps you become a true citizen of the world.⏎History of Global Diversity Awareness Month⏎After the Second World War, The Universal Declaration of Human Rights was adopted by the United Nations General Assembly at the Palais de Cha

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: how can we have strategy strategy strategy strategy strategy strategy strategy strategy strategy strategy strategy strategy strategy strategy strategy strategy strategy strategy strategy strategy strategy strategy strategy strategy strategy strategy strategy strategy strategy strategy strategy strategy strategy strategy strategy strategy strategy strategy strategy strategy strategy strategy strategy strategy

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

    = vsin(theta) Immediately after take off : vx = 2cos25 = 1.8126 vy = 2sin25 = 0.845

  2. #2 peak

    estimated as n log102. Thus, the probability is about cxn, where c = (10/9)2 ≈ 1.2 and x = (9/10)log102 ≈

  3. #3 peak

    is intended.⏎Btu (British Thermal Unit) — The amount of heat required to raise the temperature of one pound of water one degree Fahrenheit; equal to 252 calories.⏎Charge Controller — A component of a phot

  4. #4 peak

    the satellite view is shown instead.⏎Latitude: 52.5162 / 52°30'58"N⏎Longitude: -3.0604 / 3°3'37"W⏎OS

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. L26:1564 strongly suppresses
  2. L19:753 excites
  3. L8:10 suppresses
  4. L8:14 suppresses
  5. L8:16 excites
  6. L20:312 excites
  7. L8:7 suppresses
  8. L18:569 suppresses
  9. L24:226 excites
  10. L6:13 suppresses
  11. L25:11 excites
  12. L22:201 suppresses
  13. L8:4 suppresses
  14. L1:2677 excites
  15. L8:13 excites
  16. L6:8 excites
  17. L8:3 suppresses
  18. L7:62 suppresses
  19. L6:18 suppresses
  20. L5:216 excites
  21. L18:576 excites
  22. L6:12 suppresses
  23. L22:225 excites
  24. L20:52 excites

Feeds later layers

None: this is the last layer.

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 -233.63 and -225.731 values between -225.73 and -217.838 values between -217.83 and -209.946 values between -209.94 and -202.0414 values between -202.04 and -194.1533 values between -194.15 and -186.2569 values between -186.25 and -178.3699 values between -178.36 and -170.46125 values between -170.46 and -162.57175 values between -162.57 and -154.67234 values between -154.67 and -146.78256 values between -146.78 and -138.88251 values between -138.88 and -130.99280 values between -130.99 and -123.09255 values between -123.09 and -115.20276 values between -115.20 and -107.30269 values between -107.30 and -99.41241 values between -99.41 and -91.51201 values between -91.51 and -83.61193 values between -83.61 and -75.72138 values between -75.72 and -67.82134 values between -67.82 and -59.93105 values between -59.93 and -52.0399 values between -52.03 and -44.1465 values between -44.14 and -36.2470 values between -36.24 and -28.3557 values between -28.35 and -20.4584 values between -20.45 and -12.5643 values between -12.56 and -4.6641 values between -4.66 and 3.2334 values between 3.23 and 11.1323 values between 11.13 and 19.0312 values between 19.03 and 26.9210 values between 26.92 and 34.823 values between 34.82 and 42.711 values between 42.71 and 50.611 values between 58.50 and 66.401 values between 74.29 and 82.19offnegativepositiveoften

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. urers
  2. ellers
  3. Malays
  4. USERNAME
  5. ocache
  6. MMC
  7. TTC
  8. .PERMISSION
  9. antal
  10. кли

Suppresses

  1. โอกา
  2. đự
  3. וסיף
  4. nawet
  5. ในฐานะ
  6. 旗舰店
  7. ,...⏎⏎
  8. ım
  9.    
  10. +,