Neuron Explorer

Qwen3-0.6B

middle of personal names

A neuron in layer 19 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 the middle pieces of people's names, such as the 'ima' in Ibrahima or the 'i' in Devi. Turning it up made the model talk about 'the outside world' or wander off into unrelated rambling, so the steering shows no 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

    al. Take, for example, the crew of the 69-foot-long Mansor Sakho (she is painted sky blue). There is Ibrahima from Gambia. Fili from Dakar. Issa the Fulani,

  2. #2 peak

    villages of Ciucan and Tuta to give eye and dental tests.⏎For 14-year-old Tuta resident Annamaria Laura Lupo, the eye exam was long overdue. She said she's been having

  3. #3 peak

    The Geological Survey of India has identified granite waste as the construction material for groynes along vulnerable stretches of the Tamil Nadu coast.⏎GSI Deputy Director-General S.K. Bhushan told a conference on environmental geosciences organised by Anna University's

  4. #4 peak

    was born on 23 January, 1897 in Cuttack (Orissa) to Janakinath Bose and Prabhavati Devi. Janakinath Bose was one of the successful lawyer in Cuttack and

  5. #5 peak

    assassination of Lt Col NS Simpson, the infamous Inspector General of Prisons. Three Bengali revolutionaries – Benoy Basu, Badal Gupta and Dinesh Gupta – disguised themselves as Westerners to get inside the Writers’ Building

  6. #6 peak

    slavery. The woman, Hadijatou Mani Koraou, had been sold into slavery as a child and then forced to

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: It's not just the person who is in the room, but the space around them. But how to make it possible to achieve the meaning of the space, and the meaning of the person, without any prior information about the person or the

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

    and controlled by an adjustable valve and a scaven<|endoftext|>Deutsch: Faktor / Español: Factor - Factor / Français: Facteur - Facteur / Italiano: Fattori - F

  2. #2 peak

    are designed to hold the workpiece in a specific position and orientation, allowing for consistent and accurate processing.⏎Deutsch: Frequenz / Español: Frecuencia / Português: Frequéncia / Français: Frêquence / Italiano

  3. #3 peak

    , so the period is the reciprocal of the frequency.⏎Deutsch: Fabrik / Español: Fábrica / Português: Fábrica / Français: Usine / Italiano: Fabbrica⏎A factory (pre

  4. #4 peak

    where workers manufacture goods or operate machines processing one product into another.⏎Deutsch: Fell / Español: Pelaje / Français: Fourrure⏎Fur is a synonym for hair, used in

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. L18:1345 suppresses
  2. L18:1099 excites
  3. L14:1550 suppresses
  4. L18:1861 excites
  5. L18:2361 suppresses
  6. L18:2289 excites
  7. L18:269 suppresses
  8. L12:1276 suppresses
  9. L15:2128 suppresses
  10. L18:2513 suppresses
  11. L17:1257 suppresses
  12. L17:1387 excites
  13. L14:1796 suppresses
  14. L18:1199 excites
  15. L18:2225 suppresses
  16. L18:1536 excites end of web page disclaimer
  17. L18:2926 excites
  18. L18:891 suppresses
  19. L18:2732 excites
  20. L18:1477 weakly suppresses
  21. L18:1224 weakly suppresses
  22. L18:2539 weakly excites
  23. L17:2956 weakly suppresses
  24. L17:2726 weakly suppresses

Feeds later layers

  1. L22:2433 excites
  2. L20:551 excites
  3. L20:1038 excites
  4. L20:1377 excites
  5. L20:2345 weakly excites
  6. L20:1051 weakly excites
  7. L25:2226 weakly excites
  8. L20:573 weakly excites
  9. L21:627 weakly excites
  10. L20:3053 weakly suppresses
  11. L21:465 weakly suppresses
  12. L21:1754 weakly excites 'in' before a place name
  13. L23:2252 weakly excites
  14. L26:1368 weakly excites
  15. L25:1228 weakly excites
  16. L22:133 weakly suppresses
  17. L21:582 weakly excites
  18. L22:2278 weakly excites
  19. L23:2408 weakly excites
  20. L22:2317 weakly excites
  21. L25:98 weakly suppresses
  22. L24:2719 weakly excites
  23. L22:2062 weakly excites
  24. L20:1840 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 -28.31 and -27.311 values between -26.31 and -25.301 values between -20.29 and -19.291 values between -18.29 and -17.281 values between -17.28 and -16.281 values between -14.27 and -13.274 values between -13.27 and -12.271 values between -11.27 and -10.266 values between -9.26 and -8.265 values between -8.26 and -7.2612 values between -7.26 and -6.2511 values between -6.25 and -5.2517 values between -5.25 and -4.2525 values between -4.25 and -3.2433 values between -3.24 and -2.2477 values between -2.24 and -1.24743 values between -1.24 and -0.242902 values between -0.24 and 0.7738 values between 0.77 and 1.7716 values between 1.77 and 2.777 values between 2.77 and 3.782 values between 3.78 and 4.782 values between 5.78 and 6.781 values between 10.79 and 11.80offnegativepositiveoften

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. %p
  2. Oversight
  3. %f
  4. lemetry
  5. vette
  6. eva
  7. ostringstream
  8. %c
  9. %H
  10. rowave

Suppresses

  1. ечно
  2. ::⏎
  3. iali
  4. Cla
  5. Cri
  6. ести
  7. 空间
  8. cla
  9. ENTIAL
  10. Joi