Neuron Explorer

Qwen3-0.6B

unclear: 'High' in school names (negative)

A neuron in layer 5 of Qwen3-0.6B. Its strongest snippets are fairly consistent: several of them peak on the same token. It is featured, so it has steering outputs.

What this neuron does

This neuron is hard to pin down. Its strongest dips come from the word 'High' in school names like 'Bemidji High School', while its positive peaks are scattered across unrelated words such as 'Urban' in 'Housing and Urban Development', 'canton' and 'native'. Turning it up mostly made the text wander off topic, so no clear everyday theme shows up.

Written by Claude from this neuron's activations and steering outputs. Confidence: low. 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

    .<|endoftext|>Did you know that at superspeedways like Daytona, two or more cars running nose-to-tail can run faster than one?⏎On the racetrack, both drag and downforce are affected by the air flow off nearby cars.

  2. #2 peak

    : Mark D. Shroder⏎- Associate Editor: Michelle P. Matuga⏎Measuring Housing Affordability⏎U.S. Department of Housing and Urban Development⏎The views expressed in this article are those of the author and do not

  3. #3 peak

    and patriotism? This becomes a particularly good question, if not a surprising one, because, historically speaking, ideas of nationalism and patriotism have never emerged from military cantonments, the soldier’s barracks or from a general’s writings. If anything,

  4. #4 peak

    as you will often notice, there are some basic differences and similarities in each of these cases and it comes down to a number of language writing and speaking rules.⏎Native English Hacks and techniques for foreign students⏎A desire to start writing English like

  5. #5 peak

    2009 (for more, read below); and,Laikipia region of Kenya (January, 2009), a German canton of Switzerland (March, 2009), and the Piemonte

  6. #6 peak

    of combining content and language learning (CLIL) comes from countries such as the USA and Australia, where many immigrant people are forced to learn English as the second language, so that, they will be accepted in the society. In addition, it

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 city has a lot of power to the point of the city's highest education system, and the college has a lot of power to the point of the city's highest education system. The city's highest education system has a

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

    's population as you watch the video program.⏎(This activity is based in part on a lesson by Artis West of North Cobb High School in Kennesaw, Georgia. That lesson plan is included in Featured Lesson

  2. #2 peak

    The World As It Could Be Tree Dedication Ceremony, students, teachers, and other collaborators on the project reflect on the Universal Declaration of Human Rights. Balboa High School Tree Ceremony from BAYCAT.<|endoftext|>Several images have been returned to Earth

  3. #3 peak

    Nancy Riestenberg, School Climate Specialist at the Department of Education.⏎Riestenberg joined members of the task force to speak with students from Bemidji High School and students from Schoolcraft to discuss their concerns about bullying and ways in which

  4. #4 peak

    ujillo grew up in Corcoran, CA and was one of seven children in a family who earned a living as farm workers. While at Corcoran High School,

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. L2:1564 excites
  2. L4:2822 excites
  3. L3:1950 excites
  4. L3:1244 excites
  5. L3:1216 weakly suppresses
  6. L4:182 weakly suppresses
  7. L4:23 weakly excites
  8. L4:215 weakly suppresses
  9. L3:1437 weakly excites
  10. L4:25 weakly excites
  11. L4:19 weakly suppresses
  12. L4:2927 weakly suppresses
  13. L4:21 weakly excites
  14. L3:684 weakly suppresses
  15. L3:2434 weakly excites
  16. L3:917 weakly suppresses
  17. L3:1831 weakly excites
  18. L1:2782 weakly excites
  19. L2:627 weakly excites
  20. L1:1111 weakly suppresses
  21. L4:1818 weakly excites
  22. L4:989 weakly suppresses
  23. L4:2123 weakly excites
  24. L3:1889 weakly excites

Feeds later layers

  1. L6:719 suppresses
  2. L7:724 excites
  3. L7:272 suppresses
  4. L8:265 suppresses
  5. L10:2102 suppresses
  6. L6:307 suppresses
  7. L10:404 excites
  8. L11:598 excites
  9. L7:845 suppresses
  10. L10:2438 suppresses
  11. L9:181 suppresses
  12. L7:267 weakly suppresses
  13. L7:1054 weakly excites
  14. L15:1012 weakly suppresses
  15. L6:1639 weakly excites
  16. L15:715 weakly suppresses
  17. L8:2137 weakly suppresses
  18. L7:1196 weakly suppresses
  19. L9:1341 weakly suppresses
  20. L24:671 weakly suppresses
  21. L10:892 weakly excites
  22. L12:1184 weakly suppresses
  23. L13:2360 weakly suppresses
  24. L16:2835 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 -7.84 and -7.571 values between -5.69 and -5.421 values between -4.07 and -3.811 values between -3.27 and -3.001 values between -2.19 and -1.921 values between -1.92 and -1.662 values between -1.66 and -1.393 values between -1.12 and -0.855 values between -0.85 and -0.5871 values between -0.58 and -0.311614 values between -0.31 and -0.042070 values between -0.04 and 0.23112 values between 0.23 and 0.5011 values between 0.50 and 0.767 values between 0.76 and 1.033 values between 1.03 and 1.301 values between 1.30 and 1.571 values between 1.57 and 1.841 values between 2.11 and 2.381 values between 2.65 and 2.91offnegativepositiveoften

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. icing
  2. onet
  3. 也有一些
  4. тор
  5. ront
  6. jednoc
  7. 一个是
  8. rende
  9. enge
  10. Di

Suppresses

  1. Var
  2. arlo
  3. 产业集聚
  4. atters
  5. Var
  6. -dismiss
  7. ndata
  8. houses
  9. Pvt
  10. embre