Neuron Explorer

Qwen3-0.6B

"latter" and "former" referring back

A neuron in layer 6 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 "latter" (or "former") when it points back to the second (or first) of two things just mentioned. Turning it up made the model's text drift into loose, rambling story-like sentences, which doesn't clearly relate to the trigger.

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

    : should he send the boys off to jail forthwith, or should he see whether they might not be able to straighten themselves out? He decided on the latter course, and in looking for a means of rehabilitation he selected an unusual instrument:

  2. #2 peak

    for crop yields by all four methods. Overall, global yields increase ~10% under the three SGs and decrease 5% under emissions reduction, the latter primarily due to reduced CO2 fertilization, relative to business as usual by the

  3. #3 peak

    primarily as a nutritive process2. Health and illness, in this commonsensical view, are functions respectively of optimum nutrition and the lack of it. The latter could have two forms – over nutrition and under nutrition. This two-fold classification of

  4. #4 peak

    plumage with a black back.⏎- ‘Neophytes at duck identification often confuse a bufflehead with a male hooded merganser when the latter's crest is raised showing a large patch of white feathers.’⏎- ‘These

  5. #5 peak

    opportunities may also be associated with the use of carrier multiplication in solar-fuel technologies and specifically, the production of hydrogen by photo-catalytic water splitting. The latter process requires four electrons per water molecule and its efficiency can be dramatically enhanced if these

  6. #6 peak

    the expansive power of the State and society over individuals? Not always successful, and fraught with problems, but relevant enough for an intellectual tradition to be discovered. The latter can be defended, but the former demands us to jettison its nationalistic

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 need to get this information to the students. This is a great way to show what you really want to learn and share what you care about the students, and the final presentation to the students of the class, so you can use the

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

    Top, and William poured volleys of buckshot upon the enemy, then charged them with his bayonet. He recollected that one of the last shots of Gettysburg was fired at him.⏎“Next morning

  2. #2 peak

    Picchu, the mountain behind Machu Picchu. Alternatively, climb to the top of the less crowded Machu Picchu peak, which catches the first rays of the rising sun. Both mountains overlook the entire site, down to the river below

  3. #3 peak

    .⏎Pluto is the last one of the major planet. It's definitely different from the others. Discovered in

  4. #4 peak

    files, among them first Pdf files consists of Textbook Questions Answers whereas in second Pdf files you will get the additional extra questions for practices. In the third and last Pdf files, you will get full exercise of this chapter.⏎You

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. L3:2400 excites
  2. L5:110 suppresses
  3. L4:2665 weakly excites
  4. L5:1698 weakly excites
  5. L0:1672 weakly suppresses
  6. L5:169 weakly excites
  7. L0:60 weakly excites
  8. L5:429 weakly suppresses
  9. L5:50 weakly excites
  10. L1:2340 weakly suppresses
  11. L0:2877 weakly suppresses
  12. L5:1891 weakly suppresses
  13. L5:631 weakly excites
  14. L4:1128 weakly suppresses
  15. L4:867 weakly suppresses
  16. L1:867 weakly excites
  17. L3:1340 weakly excites
  18. L3:1772 weakly excites
  19. L2:70 weakly excites
  20. L5:1782 weakly excites
  21. L4:850 weakly suppresses
  22. L4:2808 weakly excites
  23. L5:1195 weakly excites
  24. L0:1182 weakly excites

Feeds later layers

  1. L7:902 excites
  2. L9:1101 weakly excites
  3. L7:559 weakly excites
  4. L13:2489 weakly suppresses
  5. L8:2806 weakly excites
  6. L8:858 weakly excites
  7. L19:1260 weakly excites
  8. L8:901 weakly excites
  9. L8:1542 weakly suppresses
  10. L19:154 weakly excites
  11. L25:522 weakly suppresses
  12. L21:2754 weakly suppresses
  13. L8:1369 weakly suppresses
  14. L19:2058 weakly suppresses
  15. L13:1286 weakly suppresses
  16. L8:2636 weakly excites
  17. L13:1040 weakly excites
  18. L10:2980 weakly suppresses
  19. L16:2091 weakly suppresses
  20. L12:2906 weakly suppresses
  21. L8:1563 weakly excites
  22. L17:2322 weakly suppresses
  23. L14:2124 weakly suppresses
  24. L12:2576 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 -1.94 and -1.881 values between -1.50 and -1.441 values between -1.01 and -0.951 values between -0.95 and -0.882 values between -0.57 and -0.516 values between -0.51 and -0.453 values between -0.45 and -0.3921 values between -0.39 and -0.3250 values between -0.32 and -0.26113 values between -0.26 and -0.20240 values between -0.20 and -0.14482 values between -0.14 and -0.08760 values between -0.08 and -0.01937 values between -0.01 and 0.05581 values between 0.05 and 0.11358 values between 0.11 and 0.17192 values between 0.17 and 0.2481 values between 0.24 and 0.3040 values between 0.30 and 0.3621 values between 0.36 and 0.428 values between 0.42 and 0.489 values between 0.48 and 0.55offnegativepositiveoften

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. Yourself
  2. matrimon
  3. ye
  4. yourself
  5. [max
  6. yyyy
  7. -scrollbar
  8. yy
  9. INESS
  10. tamil

Suppresses

  1. inos
  2. 利亚
  3. ulos
  4. 能看到
  5. aina
  6. entials
  7. %",
  8. 但他们
  9. voir
  10. %",⏎