Neuron Explorer

Qwen3-0.6B

'Pr' starting capitalised names

A neuron in layer 26 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

It fires on the 'Pr' at the start of a capitalised name, most clearly in 'Pravda', 'Prussian', 'Prinz' and 'Prudhoe Bay'. Turning it up did not make the model write anything about these names, and the outputs wandered into unrelated 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

    and bread; All power to the soviets.⏎He was caught up in a struggle in the Bolshevik party and called a betrayer of socialism because of the Pravda publishing support for the war effort and the provisional gouvernment.⏎He

  2. #2 peak

    ast, Russia), also called "the scattered church of Prussian Lithuania." The Mennonites living here had much to endure from the dragoons of the Prussian King Friedrich Wilhelm I, who finally expelled them from the district. The [[

  3. #3 peak

    ideal platform for this work.⏎Angela Prinz from Germany is the team leader. She confirms that Eunice had a

  4. #4 peak

    places of Polish Prussia.⏎The Dutch Committee of Foreign Needs organized a general offering in May 1

  5. #5 peak

    parts and the diagrams found in a computer magazine. It was Wennberg who came up with the man to do just that. After discussion with another friend, Wes Prichard, Prichard suggested to Wennberg<|endoftext|>Home > Preview⏎The

  6. #6 peak

    overthrow of the Romanov Dynasty⏎Role of Stalin in the 1917 Revolution⏎He was released from exile and became part of the editorial board of Pravda.⏎He was elected to the

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 can an untrained staff member which seenoach seenoach seenoach seenoach seenoach seenoach seenoach seenoach seenoach seenoach seenoach seenoach seenoach se

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

    cers, acid reflux, and gastritis. The mucilage of the plant’s pulp has restorative and protective qualities to the stomach lining, and applied externally to sprains, burns, and bruises reduce swelling, discoloration and inflammation while speeding

  2. #2 peak

    maintain healthy blood sugar levels. Ginger, another key spice in pumpkin pie, is well-known for its benefits as a digestive aid and is particularly helpful for nausea. Cloves can benefit digestion and the immune system, and nutmeg can

  3. #3 peak

    of yellowfin tuna (Thunnus albacares) in nearshore areas in French Polynesia, determined using ultrasonic telemetry⏎Centre IRD de Montpellier, BP 5045, 34032 Mont

  4. #4 peak

    RNA from chemically complex plants⏎More information: The draft genome of the carcinogenic human liver fluke Clonorchis sinensis, Xiaoyun Wang, Wenjun Chen, Yan

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. L25:741 strongly excites
  2. L22:390 excites
  3. L24:1707 suppresses
  4. L24:2239 excites
  5. L23:493 excites
  6. L24:25 suppresses
  7. L20:692 suppresses
  8. L18:2989 weakly excites
  9. L18:345 weakly excites
  10. L24:242 weakly excites
  11. L24:1421 weakly suppresses
  12. L20:2494 weakly excites
  13. L7:122 weakly excites
  14. L8:7 weakly excites
  15. L13:2993 weakly excites
  16. L18:569 weakly excites
  17. L9:2440 weakly excites
  18. L4:784 weakly suppresses
  19. L19:2358 weakly excites "any of" before "these" or "the"
  20. L19:353 weakly excites
  21. L12:2167 weakly suppresses
  22. L15:2621 weakly suppresses
  23. L20:2581 weakly excites
  24. L12:469 weakly excites

Feeds later layers

  1. L27:2782 suppresses
  2. L27:467 suppresses
  3. L27:713 suppresses
  4. L27:239 excites
  5. L27:587 weakly excites
  6. L27:1266 weakly excites
  7. L27:1807 weakly suppresses
  8. L27:1584 weakly suppresses
  9. L27:2957 weakly excites
  10. L27:46 weakly excites
  11. L27:2841 weakly suppresses
  12. L27:2988 weakly suppresses capital M starting a word
  13. L27:988 weakly excites word before 'needed' or 'required'
  14. L27:2034 weakly suppresses
  15. L27:1155 weakly suppresses just before a pronoun for God
  16. L27:977 weakly excites
  17. L27:2111 weakly excites
  18. L27:1947 weakly suppresses
  19. L27:72 weakly excites
  20. L27:775 weakly suppresses
  21. L27:1644 weakly excites
  22. L27:1757 weakly excites
  23. L27:1116 weakly excites
  24. L27:1880 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 -8.82 and -8.161 values between -7.50 and -6.842 values between -6.84 and -6.171 values between -6.17 and -5.511 values between -5.51 and -4.852 values between -4.85 and -4.191 values between -4.19 and -3.532 values between -3.53 and -2.873 values between -2.87 and -2.217 values between -2.21 and -1.5410 values between -1.54 and -0.88257 values between -0.88 and -0.223551 values between -0.22 and 0.4449 values between 0.44 and 1.1010 values between 1.10 and 1.763 values between 1.76 and 2.431 values between 2.43 and 3.091 values between 3.09 and 3.751 values between 3.75 and 4.411 values between 8.38 and 9.041 values between 9.70 and 10.361 values between 11.03 and 11.691 values between 16.98 and 17.64offnegativepositiveoften

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. };⏎⏎⏎
  2. зон
  3. '];?></
  4. };⏎⏎⏎
  5. };⏎⏎⏎⏎
  6. };⏎
  7. '};⏎
  8. }; ⏎ ⏎
  9. };⏎⏎
  10. };⏎

Suppresses

  1. ())
  2. ) ⏎
  3. ")
  4. 另一半
  5. ()>⏎
  6. ")
  7. ") ⏎
  8. )⏎
  9. &)
  10. "]