Neuron Explorer

Qwen3-0.6B

years and numbers before the digits

A neuron in layer 25 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 small word just before a year or other number, such as the space before '1712' in 'in 1712' or before '12th Century'. When it was turned up, the text did not clearly change in any readable theme, so its job is hard to pin down from the outputs.

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

    is not just a hugely significant place of worship, but a gigantic repository of architectural and artistic splendour. There has been a church on this site since around 627 AD, when a wooden structure was erected over some part of the

  2. #2 peak

    sale.⏎The term originated with the cistercian monks at Eberbach Abbey in Rheingau, where the first recorded use of the term occurred in 1712. The abbey’s best wines were set aside to be

  3. #3 peak

    clues as to what life was like in medieval Edinburgh,” he said.⏎Excavation work has already revealed the precise location of the monastery, which was founded in 1230 by King Alexander II, the ruler of Scotland from 1

  4. #4 peak

    arrived in Scotland after the Norman conquest of England in 1066. There is a record of a David de Boivil witnessing a charter as early as 1164.<|endoftext|>For the questions below, use the most efficient Excel

  5. #5 peak

    the use of black and thick wooden. Most Gothic glass windows have large, pointed, Gothic-styled windows.⏎Gothic architecture has been around since typically the 12th Century. This was founded upon the idea that architecture need to be

  6. #6 peak

    we read:⏎Ebionites. This designation was at first like 'Nazarenes,' a common name for all Christians, as Epiphanius (d. 403) testifies (Adv. Har. xxix.1

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 to align the learning stage with the stage of learning and align learning stages properly with learning stages of the person's growing stage phase, what stage stage stage stage stage stage stage stage stage stage stage stage stage stage stage stage stage stage stage stage

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

    8⏎Ben sold 830 newspapers. His brother James sold 177 fewer newspapers than Ben.⏎(a) How many newspapers did James sell?⏎(b) How many newspapers did they sell altogether?⏎In early 2-step problems,

  2. #2 peak

    c) is called programmable array logic , first introduced by Monolithic Memories as the PAL series of devices.⏎Because the arrangement of Figure 5.11 (c) is very similar to a ROM, we sometimes call a horizontal product-term

  3. #3 peak

    the tables representing your daily salary for your first two weeks at work, assuming you take the job.⏎b. Write an equation to model the growth of your salary.⏎c. How much will you

  4. #4 peak

    construct the product terms) that feed OR gates (which form the sums of the product terms). We can simplify this representation still further ( Figure 5.11 c), by drawing the input lines to a multiple-input AND gate as if they

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. L24:485 suppresses
  2. L23:333 suppresses
  3. L24:658 excites
  4. L23:2220 excites
  5. L23:2334 excites
  6. L22:1984 suppresses unclear: word before a count or quantity
  7. L6:1701 weakly suppresses
  8. L22:672 weakly suppresses
  9. L16:3071 weakly suppresses
  10. L7:1017 weakly excites
  11. L10:1304 weakly suppresses
  12. L24:2080 weakly excites
  13. L23:2364 weakly excites
  14. L1:984 weakly suppresses
  15. L10:858 weakly excites digit 1 in COVID-19 and 1941
  16. L21:2932 weakly suppresses
  17. L9:2422 weakly suppresses
  18. L7:466 weakly excites
  19. L6:1775 weakly suppresses
  20. L22:2 weakly excites
  21. L2:2910 weakly suppresses
  22. L18:1315 weakly suppresses
  23. L9:2480 weakly suppresses
  24. L0:1012 weakly suppresses "of" after nouns like eve, family, modes

Feeds later layers

  1. L26:1055 strongly excites
  2. L26:546 strongly excites
  3. L27:495 strongly excites
  4. L26:297 strongly suppresses
  5. L26:1427 excites
  6. L26:519 excites
  7. L27:779 weakly excites
  8. L27:629 weakly excites
  9. L27:18 weakly suppresses
  10. L27:668 weakly excites
  11. L26:1353 weakly excites
  12. L26:2543 weakly suppresses
  13. L26:1942 weakly suppresses
  14. L27:3012 weakly suppresses
  15. L26:2859 weakly suppresses
  16. L27:2710 weakly suppresses
  17. L26:2019 weakly excites
  18. L26:1178 weakly excites
  19. L27:2382 weakly suppresses
  20. L26:935 weakly suppresses
  21. L27:2603 weakly excites
  22. L26:1038 weakly suppresses
  23. L26:1296 weakly suppresses word-start 'mult' fragment
  24. L27:940 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 -11.84 and -10.721 values between -7.39 and -6.272 values between -6.27 and -5.166 values between -5.16 and -4.0511 values between -4.05 and -2.9412 values between -2.94 and -1.83105 values between -1.83 and -0.713310 values between -0.71 and 0.40340 values between 0.40 and 1.5157 values between 1.51 and 2.6215 values between 2.62 and 3.748 values between 3.74 and 4.8511 values between 4.85 and 5.967 values between 5.96 and 7.076 values between 7.07 and 8.194 values between 8.19 and 9.303 values between 9.30 and 10.412 values between 10.41 and 11.522 values between 11.52 and 12.641 values between 15.97 and 17.081 values between 18.20 and 19.311 values between 24.87 and 25.981 values between 27.10 and 28.211 values between 31.54 and 32.66offnegativepositiveoften

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. c
  2. c
  3. (c
  4. *c
  5. )c
  6. =c
  7. #c
  8. [c
  9. ,c
  10. $c

Suppresses

  1. Consolid
  2. 铉
  3. 党组书记
  4. работка
  5. ไกล
  6. ┬
  7. ฉบ
  8. раф
  9. lửa
  10. нд