Neuron Explorer

Qwen3-0.6B

'Indian' before Ocean or subcontinent

A neuron in layer 24 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 reacts to the word "Indian" when it comes before "Ocean", "subcontinent" or a place name, such as "the Indian Ocean". Turning it up or down did not give a clear theme in the generated text.

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

    Rasikajy” occupied settlements along the coast. They were exploiting the natural resources of the rich hinterland and participating in the large trade network of the Indian Ocean.⏎New fieldwork research is ongoing to investigate the production of stone vessels and

  2. #2 peak

    Suez Canal⏎The Suez Canal connects the Mediterranean Sea (foreground left) with the Gulf of Suez and thereby the Red Sea and the Indian Ocean and this way shortens the sea passage from Europe to the Middle and Far

  3. #3 peak

    faculty position in Computer Science at Queen’s University Belfast, United Kingdom. He received his B.Tech from CUSAT and his M.Tech and PhD from Indian Institute of Technology Madras, all in Computer Science. His current research interests include

  4. #4 peak

    AnimalWised and discover the native animals of the Thar Desert.⏎The Thar Desert: Home to diverse species⏎Located in the northwestern region of the Indian subcontinent, between Pakistan and India, the Thar desert has minimal vegetation yet

  5. #5 peak

    Hawai‘i: Northwest Hawaiian⏎Islands, O‘ahu, Maui, and Hawai‘i Island.⏎Worldwide: Eastern Atlantic,⏎Caribbean, Indian and Pacific Oceans.⏎Before the introduction of⏎Acanthophora sp

  6. #6 peak

    Viswanathan said with increasing degradation of the earth, environmental geosciences assumed critical importance. The recent earthquakes in Sumatra and Kashmir and the tsunami in the Indian Ocean had created greater awareness about the importance of environmental geosciences among policy makers

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 want to add a feature to a web app that is full of small black tiles that look like a black tile URL URL URL URL URL URL URL URL URL URL URL URL URL URL URL URL URL URL URL URL URL URL URL URL URL

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

    it is freed from the clutches of British rule. During the partition of Bengal and freedom by the indigenous movement in which Sir Henry Cotton Surendra Nath Banerjee and turned into a

  2. #2 peak

    not be a recovery at all, but just an L-shaped continuation of economic misery.)⏎But not everyone subscribes to this framework of thought. Anirvan Banerji, director of research at

  3. #3 peak

    haMbhAraveNa cha |⏎vrajasyotthApanaM chakrurghoShavR^iddhAH samAgatAH ||2-9<|endoftext|>A population of fruit bats which is found across much

  4. #4 peak

    Dakini⏎- The Buddhist protector deity Penden Lhamo⏎- The Goddess of the White Umbrella (the Buddhist deity Ushnisha-sitatapatra)⏎- The Great mystic Virupa⏎- Thunderbolt and bell⏎

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. L23:397 strongly excites
  2. L20:500 strongly suppresses
  3. L19:1228 excites
  4. L10:190 excites
  5. L10:2008 suppresses
  6. L21:2222 suppresses
  7. L20:1027 excites
  8. L18:477 excites
  9. L9:2483 excites
  10. L18:177 suppresses
  11. L3:62 excites
  12. L16:692 suppresses
  13. L9:144 suppresses
  14. L18:560 suppresses
  15. L10:556 excites
  16. L4:1890 suppresses
  17. L9:429 excites
  18. L7:546 suppresses
  19. L8:495 suppresses
  20. L20:2345 excites
  21. L9:812 excites
  22. L9:2482 suppresses
  23. L11:840 suppresses
  24. L21:285 excites

Feeds later layers

  1. L27:2458 excites
  2. L25:1504 weakly excites
  3. L25:2687 weakly excites
  4. L25:1525 weakly excites
  5. L25:2761 weakly suppresses
  6. L27:847 weakly excites
  7. L26:2807 weakly excites
  8. L26:2004 weakly excites
  9. L25:2086 weakly suppresses
  10. L25:574 weakly excites
  11. L25:1804 weakly suppresses
  12. L25:254 weakly excites
  13. L25:2307 weakly suppresses
  14. L25:485 weakly excites
  15. L25:2725 weakly excites
  16. L25:2432 weakly excites
  17. L27:1001 weakly excites
  18. L26:1991 weakly excites
  19. L25:1078 weakly suppresses
  20. L25:2203 weakly excites
  21. L25:230 weakly suppresses
  22. L25:2795 weakly excites
  23. L25:2458 weakly excites
  24. L27:516 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 -13.41 and -12.471 values between -12.47 and -11.531 values between -8.71 and -7.773 values between -6.83 and -5.894 values between -5.89 and -4.953 values between -4.95 and -4.012 values between -4.01 and -3.076 values between -3.07 and -2.1350 values between -2.13 and -1.191188 values between -1.19 and -0.252497 values between -0.25 and 0.6989 values between 0.69 and 1.6326 values between 1.63 and 2.5711 values between 2.57 and 3.514 values between 3.51 and 4.455 values between 4.45 and 5.392 values between 5.39 and 6.332 values between 7.27 and 8.212 values between 8.21 and 9.152 values between 11.97 and 12.913 values between 12.91 and 13.852 values between 14.79 and 15.731 values between 17.61 and 18.551 values between 18.55 and 19.491 values between 21.37 and 22.311 values between 23.25 and 24.19offnegativepositiveoften

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. _IND
  2. inds
  3. Ind
  4. IND
  5. _ind
  6. Indians
  7. Ind
  8. Gandhi
  9. IND
  10. _definition

Suppresses

  1. เกร
  2. reibung
  3. epub
  4. zo
  5. spots
  6. Appe
  7. 圩
  8. bug
  9. خلا
  10. ник