Neuron Explorer

Qwen3-0.6B

author surnames in citations

A neuron in layer 1 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 end of a researcher's surname in a citation, such as 'Postman' or 'Pullman', and on 'et' in 'et al.'. It turns strongly negative on the word 'of' right after a researcher's name. Steering gave no consistent topic.

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

    strongly reflects the view advocated by Neil Postman (1979) of creating a “thermostatic mechanism” that

  2. #2 peak

    lpd.⏎How to pronounce daemon His Dark Materials?⏎a demon (/ˈdiːmən/listen)) is a type of fictional being in Philip Pullman's fantasy trilogy His Dark Materials.⏎How do you say mail daemon?⏎So the

  3. #3 peak

    what happens when Lila meets Razvani and if she survives to tell the tale.⏎I think we need to look out for some more stories by Philip Pullman. Can you recommend any for us?⏎Year 4 Blog⏎We are Miss

  4. #4 peak

    Quintín Bandera Betancourt (ca. 1834 – 1906), also Quentin Bandera and Quintin Bandera, was a military leader of the Cuban insurrection against the Spanish during the Cuban War of

  5. #5 peak

    Economics,⏎American Economic Association, vol. 3(2), pages 137-66, April.⏎- Angus Deaton & Olivier Dupriez, 2009. "Purch<|endoftext|>Of Biochemistry and Bel

  6. #6 peak

    a coalition of health organizations published new guidelines for cervical cancer screening. DCEG research into the benefits of incorporating HPV testing into cervical cancer screening programs (Katki et al., 2011) have informed these revised guidelines. More information

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, and I have been thinking about lately. I have been thinking about lately, and I have been thinking about lately. I have been thinking about lately, and I have been thinking about lately. I have been

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

    because we can extract it from bones of woolly mammoths, which date back tens of thousands of years, and make sense of it," explains Nick Goldman of EMBL-EBI. "It's also incredibly small, dense and

  2. #2 peak

    the way players earn forgiveness are not necessarily congruent with harsh punishments.⏎Research by Michael Wenzel of Flinders University in Australia has illuminated the difference between two conceptions of justice.

  3. #3 peak

    or pleasure, or even sees that pumpkin bread with chocolate chips bringing someone else happiness, then it can improve a sense of wellbeing,” says researcher Jacqueline Gollan of

  4. #4 peak

    ia in human culture⏎In the last quarter of the 18th century, Professor Jacob Seydelmann of Dresden developed a process to extract and produce a more concentrated form for use in water

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. L0:1301 suppresses
  2. L0:351 suppresses
  3. L0:1224 suppresses
  4. L0:178 suppresses
  5. L0:638 weakly suppresses
  6. L0:300 weakly excites
  7. L0:454 weakly suppresses
  8. L0:1700 weakly excites
  9. L0:381 weakly excites
  10. L0:809 weakly suppresses
  11. L0:3066 weakly suppresses
  12. L0:2175 weakly excites
  13. L0:417 weakly suppresses
  14. L0:780 weakly suppresses
  15. L0:571 weakly suppresses
  16. L0:539 weakly suppresses
  17. L0:960 weakly suppresses
  18. L0:64 weakly excites
  19. L0:1534 weakly excites
  20. L0:747 weakly suppresses
  21. L0:640 weakly suppresses abbreviation periods after initials
  22. L0:577 weakly suppresses
  23. L0:1394 weakly suppresses
  24. L0:1963 weakly excites

Feeds later layers

  1. L5:1533 suppresses
  2. L2:2867 excites
  3. L7:361 weakly excites
  4. L22:2578 weakly excites
  5. L21:1175 weakly excites
  6. L3:810 weakly excites
  7. L3:334 weakly suppresses
  8. L19:1067 weakly excites
  9. L6:1430 weakly excites
  10. L15:380 weakly suppresses
  11. L2:1793 weakly excites
  12. L18:671 weakly excites
  13. L4:287 weakly suppresses
  14. L14:984 weakly suppresses
  15. L10:2918 weakly excites
  16. L8:1372 weakly excites
  17. L15:1169 weakly suppresses
  18. L9:165 weakly suppresses
  19. L21:2337 weakly excites
  20. L12:1325 weakly excites
  21. L6:1159 weakly excites
  22. L6:2616 weakly excites
  23. L6:2348 weakly excites
  24. L14:2375 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 -6.66 and -6.482 values between -4.64 and -4.461 values between -2.62 and -2.432 values between -2.07 and -1.881 values between -1.88 and -1.701 values between -1.70 and -1.514 values between -1.15 and -0.963 values between -0.96 and -0.785 values between -0.78 and -0.603 values between -0.41 and -0.2383 values between -0.23 and -0.043429 values between -0.04 and 0.14351 values between 0.14 and 0.3218 values between 0.32 and 0.514 values between 0.51 and 0.69offnegativepositiveoften

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. evity
  2. Origin
  3. ][/
  4. Favorite
  5. MethodInfo
  6. Congress
  7. ularity
  8. cerpt
  9. Covenant
  10. 无敌

Suppresses

  1. ands
  2. 和支持
  3. ポート
  4. UBLE
  5. 实验室
  6. kan
  7. andle
  8. Nd
  9. nod
  10. .Support