Neuron Explorer

Qwen3-0.6B

blank slot before a founding year

A neuron in layer 8 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 fires on the spot right after 'founded in' or 'was founded in', where a year is about to come next. Turning it up pushed the model toward unrelated text about math and weather, which does not show the same idea.

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

    Thegsun Tashi Gomang⏎The Karma Thegsun Tashi Gomang ws founded in 1980. The Gyalwa Karmapa (a Buddhist priest

  2. #2 peak

    Ballet of the Twin Cities; Immigration and Refugee Services of America; and the International Institute. Founded in 1919, the International Institute was established to provide various services for migrants

  3. #3 peak

    museum has a scientific staff of more than 200, and sponsors over 100 special field expeditions each year.⏎The Museum was founded in 1869 and housed in the old Arsenal building in Central Park. Theodore

  4. #4 peak

    the 'Hereford Herd Book Society', founded in 1878.⏎Polled Herefords were developed from the horned

  5. #5 peak

    around 1000. The monastery belonged, as does the current church, to the Benedictine Arch Abbey of Pannonhalma which was founded in 996. Around 1550 the monastery was

  6. #6 peak

    . The Society is affiliated to the European Public Health Association and is a member of the International Epidemiological Association’s European Epidemiology Federation.⏎The Society was founded in 1956.<|endoftext|>- THE MAGAZINE⏎- NEW PRODUCTS⏎

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: the fact that the present experience is a part of the current state of the previous thought. So, the present experience is a way to get to the present, and the present experience is a way to get to the present. So, 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

    use of the phrases hope, love, dwelling, college, youngsters, group, help, and assist after the drills. The researchers thought of posting with these phrases within the aftermath of the drills to be indicative of getting excessive nervousness.⏎They included

  2. #2 peak

    will be used to approximate all real zeros of a polynomial.⏎This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.⏎Spreadsheets: Laying a Foundation for Understanding Functions,⏎Spreads

  3. #3 peak

    This preview has intentionally blurred sections. Sign up to view the full version.View Full Document⏎Unformatted text preview: itself, and if ( ) ( ) lim lim x

  4. #4 peak

    red flag.⏎Copyright (c) 2022 Proceedings Of International Conference On Communication Science⏎This work is licensed under a Creative Commons Attribution 4.0 International License.<|endoftext|>How do you explain how babies are made to an 8

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. L6:72 suppresses
  2. L5:191 excites
  3. L6:1893 weakly excites
  4. L7:2752 weakly excites
  5. L7:1272 weakly suppresses
  6. L5:1363 weakly excites
  7. L7:2173 weakly suppresses
  8. L7:383 weakly excites
  9. L7:707 weakly excites
  10. L5:290 weakly excites
  11. L4:74 weakly suppresses
  12. L1:2394 weakly suppresses
  13. L1:118 weakly suppresses
  14. L0:984 weakly suppresses
  15. L4:861 weakly excites
  16. L4:1419 weakly suppresses
  17. L6:659 weakly excites
  18. L4:428 weakly excites
  19. L6:194 weakly excites
  20. L5:1455 weakly excites
  21. L4:2136 weakly excites
  22. L4:810 weakly suppresses
  23. L5:940 weakly suppresses
  24. L4:2292 weakly excites

Feeds later layers

  1. L10:2021 suppresses
  2. L14:523 suppresses
  3. L10:407 excites
  4. L14:1230 suppresses
  5. L13:929 excites
  6. L12:1791 suppresses
  7. L16:2652 suppresses
  8. L14:2360 weakly excites
  9. L14:3017 weakly suppresses
  10. L19:307 weakly suppresses
  11. L13:2528 weakly suppresses
  12. L13:1285 weakly suppresses
  13. L9:1759 weakly suppresses
  14. L9:426 weakly suppresses
  15. L16:2450 weakly suppresses
  16. L19:1431 weakly excites
  17. L10:1882 weakly suppresses
  18. L13:987 weakly suppresses
  19. L13:2262 weakly excites
  20. L15:814 weakly excites
  21. L10:1320 weakly suppresses
  22. L21:425 weakly suppresses
  23. L21:3005 weakly suppresses
  24. L17:1093 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 -1.32 and -1.261 values between -1.19 and -1.123 values between -0.99 and -0.933 values between -0.86 and -0.802 values between -0.80 and -0.732 values between -0.73 and -0.671 values between -0.67 and -0.609 values between -0.60 and -0.5311 values between -0.53 and -0.4727 values between -0.47 and -0.4048 values between -0.40 and -0.3494 values between -0.34 and -0.27153 values between -0.27 and -0.21269 values between -0.21 and -0.14399 values between -0.14 and -0.08793 values between -0.08 and -0.011023 values between -0.01 and 0.06390 values between 0.06 and 0.12223 values between 0.12 and 0.19134 values between 0.19 and 0.2596 values between 0.25 and 0.3262 values between 0.32 and 0.3852 values between 0.38 and 0.4531 values between 0.45 and 0.5125 values between 0.51 and 0.5814 values between 0.58 and 0.6511 values between 0.65 and 0.7111 values between 0.71 and 0.783 values between 0.78 and 0.844 values between 0.84 and 0.916 values between 0.91 and 0.972 values between 0.97 and 1.041 values between 1.04 and 1.101 values between 1.10 and 1.173 values between 1.23 and 1.30offnegativepositiveoften

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. 这家伙
  2. dataList
  3. itemList
  4. Analysis
  5. Analysis
  6. odelist
  7. gente
  8. ño
  9. 硌
  10. подоб

Suppresses

  1. lete
  2. ikt
  3. ermann
  4. reck
  5. amma
  6. kap
  7. лот
  8. hosts
  9. quisite
  10. gae