Neuron Explorer

Qwen3-0.6B

numbers right after 'these' or 'first'

A neuron in layer 2 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 word just before a number, especially after words like 'these', 'first', or 'those same', as in 'these 120 films'. When turned up, the model's text drifted into rambling, repeated phrases about being honest or about the news, with no clear theme.

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

    with closed-captioning. Users will also find 164 film clips (80 in English; 84 in French in the French part).⏎These 120 films are only a fraction of NFB productions on Canada’s ethnic

  2. #2 peak

    food. The various states submitted their top 10 plants to the association, and from that a list was produced for the Southeast region of the country.⏎These 10 browse species are:⏎American beautyberry (French mulberry)⏎Beg

  3. #3 peak

    and medicines.⏎By planting 20 million trees, the earth and its people will be provided with 260 million more tons of oxygen. Those same 20 million trees will remove 10 million tons of CO2

  4. #4 peak

    only 5 frogs and⏎4 worms, also to pick out at least one of each you would need to⏎pick out 15 sweets because the first 9 could be mice and the next 5⏎you pick out could be frogs

  5. #5 peak

    the conclusions of the detailed comparison analysis highlight several elements that will allow the team to refine their method and produce even more detailed maps.⏎The valuable information derived from these 2 projects forms a solid basis to support policy initiatives in Flanders as well as

  6. #6 peak

    .⏎The majority of patients (85%) fail to clear the virus within 6 months and develop chronic hepatitis C. These patients are relatively well in the first 2 decades after acquiring the infection. However in 20% of these carriers

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 am a teacher in a class. I have a class with 3 students. There are 4 students in total. I have a class with 3 students. There are 4 students in total. I have a class with

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

    of effectiveness) each year produces over $350,000 more income for his/her students compared to an average teacher. On the the other hand, a poor teacher (one in the bottom quarter in terms of effectiveness

  2. #2 peak

    Rosey the Riveter, the woman, sporting a red bandana, her bicep in the air, saying "...We Can Do It!". Or the the poster warning against spies with "...Loose Lips Sink Ships".⏎By the time you

  3. #3 peak

    , as well as nationality, were increasingly reduced to one dimension as the climate of insecurity turned nationalism into a state religion.” The all-encompassing potential for abuse within nationalism can redefine and control a people with brutal

  4. #4 peak

    to provide you, the budding programmer, the the knowledge you need on programming in an easy and fun manner. In this course I

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. L1:1874 excites
  2. L0:178 weakly suppresses
  3. L1:740 weakly suppresses
  4. L1:2341 weakly excites
  5. L0:2049 weakly excites
  6. L1:2579 weakly suppresses
  7. L0:2424 weakly suppresses
  8. L1:331 weakly suppresses
  9. L0:2989 weakly suppresses
  10. L0:193 weakly excites
  11. L1:1099 weakly excites
  12. L0:1224 weakly suppresses
  13. L1:2683 weakly excites
  14. L1:1007 weakly suppresses name pieces inside a multi-part word
  15. L1:604 weakly excites
  16. L1:1660 weakly suppresses
  17. L1:200 weakly suppresses
  18. L1:693 weakly suppresses
  19. L0:2495 weakly excites
  20. L1:124 weakly suppresses
  21. L1:2032 weakly excites
  22. L1:2593 weakly suppresses
  23. L1:1492 weakly excites
  24. L1:2569 weakly suppresses

Feeds later layers

  1. L9:537 weakly suppresses
  2. L3:1138 weakly suppresses
  3. L17:2970 weakly excites
  4. L4:159 weakly excites
  5. L7:2915 weakly suppresses
  6. L25:1435 weakly suppresses
  7. L24:1086 weakly suppresses
  8. L20:100 weakly excites
  9. L18:1547 weakly excites
  10. L23:1316 weakly suppresses
  11. L14:1020 weakly excites
  12. L23:105 weakly excites
  13. L13:2729 weakly excites
  14. L3:1538 weakly excites 'what' as a free relative
  15. L9:376 weakly suppresses
  16. L5:1194 weakly suppresses
  17. L16:1433 weakly suppresses
  18. L16:2856 weakly suppresses
  19. L18:1943 weakly suppresses
  20. L15:2762 weakly suppresses
  21. L24:303 weakly suppresses
  22. L6:2145 weakly excites
  23. L4:1851 weakly suppresses
  24. L5:2409 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 -0.89 and -0.791 values between -0.79 and -0.691 values between -0.69 and -0.595 values between -0.59 and -0.4912 values between -0.49 and -0.4031 values between -0.40 and -0.3093 values between -0.30 and -0.20434 values between -0.20 and -0.101235 values between -0.10 and -0.001271 values between -0.00 and 0.10431 values between 0.10 and 0.19243 values between 0.19 and 0.2996 values between 0.29 and 0.3923 values between 0.39 and 0.498 values between 0.49 and 0.593 values between 0.59 and 0.691 values between 0.69 and 0.781 values between 0.78 and 0.884 values between 0.98 and 1.081 values between 1.08 and 1.182 values between 1.18 and 1.282 values between 1.37 and 1.472 values between 1.47 and 1.571 values between 1.57 and 1.673 values between 1.67 and 1.771 values between 1.87 and 1.961 values between 2.06 and 2.161 values between 2.95 and 3.05offnegativepositiveoften

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. insured
  3. dots
  4. points
  5. 剂
  6. notation
  7. etes
  8. positions
  9. atorial
  10. ickets

Suppresses

  1. structuring
  2. 加载
  3. 实现
  4. reeze
  5. yên
  6. 另一个
  7. 挖掘
  8. 还将
  9. магазин
  10. 兼顾