Neuron Explorer

Qwen3-0.6B

'a' after 'in' or 'such'

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 'a' in set phrases like 'in a world', 'in a way', or 'in such a way'. It is strongest when the 'a' starts a new phrase after a preposition. Turning it up made the text wander without a 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

    ’s tricky (tricky) tricky (tricky)”⏎–Run-D.M.C., “It’s Tricky”⏎Long-term planning is hard, especially in a democracy. America’s elected regimes have short residence times, as Presidents come and go

  2. #2 peak

    ildfermentation.com/). Teaching others how to do the same is one great way of keeping traditional (and ‘living’) foods, well, alive! In a world inundated by highly processed or Genetically Modified foods that use chemicals as pres

  3. #3 peak

    ee and a, two ee and a, three ee and a, four ee and a’. I actually never learnt this way and I’ll get into that in a second.⏎By doing the above you have vocalised the 16 sixteenth

  4. #4 peak

    make life so difficult for them. Its unconventional aim is to teach students that it is O.K. to ''act autistic'' and also how to get by in a world where it is not.⏎Trained in self-advocacy, students

  5. #5 peak

    first thing that we should keep in mind while applying growth methods is that it should be properly and methodically pruned or trimmed. It should be trimmed in such a way that it encourages further growth of the plant and does not make<|endoftext|>Amaz

  6. #6 peak

    reality. A tathagata is one that has achieved awakening as to the nature of the Universe, as did Siddhartha Gotama. Being in such a state of condition is coming to know one’s own nature as is expressed in our

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 have a job to a project, and I have a budget. So I am trying to create a budget for this, I have a job to a project, and I have a budget. I am trying to create 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

    , selecting a location, glazing, environmental controls<|endoftext|>Advent Reflections: Week 2, Day 5⏎“Moses said to his father-in-law, “Because the people come to me to inquire of God. When they have

  2. #2 peak

    nevertheless had no tolerance for bad acting. Edwin bought an interest in the Winter Garden Theatre at 667 Broadway in New York City together with his brother-in-law John Sleeper Clarke. The brothers John Wilkes, Edwin, and Junius

  3. #3 peak

    faith. Deposed in the "Glorious Revolution" of 1688-89, he fled to France. His daughter and son-in-law succeeded him as Queen Mary II and King William III. James II died in

  4. #4 peak

    22. Most wanted a group of Islamic leaders to decide while a minority felt that the Prophet's brother-in-law, Ali, was the rightful successor as it should be someone from the prophet's

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. L0:1301 weakly suppresses
  2. L0:178 weakly suppresses
  3. L0:1224 weakly suppresses
  4. L1:1735 weakly excites
  5. L1:526 weakly suppresses
  6. L1:2173 weakly excites
  7. L1:2278 weakly excites
  8. L1:2801 weakly excites
  9. L1:1973 weakly suppresses
  10. L0:27 weakly excites
  11. L0:1563 weakly suppresses
  12. L0:1671 weakly suppresses
  13. L0:895 weakly excites
  14. L1:1785 weakly suppresses
  15. L1:2363 weakly excites
  16. L0:1176 weakly suppresses
  17. L1:2340 weakly suppresses
  18. L0:1644 weakly excites
  19. L1:28 weakly excites
  20. L0:2561 weakly excites
  21. L1:319 weakly suppresses "on" after effect or impact nouns
  22. L0:60 weakly excites
  23. L0:508 weakly suppresses
  24. L0:1722 weakly suppresses

Feeds later layers

  1. L11:881 excites "with" after combine, mix or react verbs
  2. L4:3008 weakly suppresses
  3. L16:1550 weakly excites
  4. L6:2562 weakly suppresses
  5. L16:1492 weakly suppresses
  6. L3:1945 weakly excites 'to' in 'person to person' phrases
  7. L11:1993 weakly excites
  8. L10:1726 weakly excites
  9. L5:2276 weakly excites
  10. L21:2551 weakly excites
  11. L14:1691 weakly suppresses
  12. L8:2017 weakly excites
  13. L22:2096 weakly excites
  14. L25:2798 weakly suppresses
  15. L18:1509 weakly suppresses
  16. L16:1978 weakly excites
  17. L5:1105 weakly suppresses
  18. L18:2425 weakly excites
  19. L18:2954 weakly suppresses
  20. L27:2514 weakly suppresses
  21. L22:839 weakly excites
  22. L18:96 weakly excites
  23. L4:2358 weakly suppresses
  24. L12:1050 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 -1.83 and -1.671 values between -1.19 and -1.033 values between -0.71 and -0.553 values between -0.55 and -0.3950 values between -0.39 and -0.23558 values between -0.23 and -0.072665 values between -0.07 and 0.09553 values between 0.09 and 0.2551 values between 0.25 and 0.416 values between 0.41 and 0.572 values between 0.57 and 0.732 values between 0.73 and 0.892 values between 0.89 and 1.051 values between 1.05 and 1.211 values between 1.85 and 2.011 values between 2.01 and 2.171 values between 2.33 and 2.491 values between 2.97 and 3.131 values between 3.13 and 3.292 values between 3.45 and 3.611 values between 3.61 and 3.771 values between 4.09 and 4.251 values between 4.41 and 4.57offnegativepositiveoften

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. scious
  2. 工业大学
  3. correlated
  4. relat
  5. tagged
  6. Against
  7. تعاون
  8. har
  9. over
  10. änner

Suppresses

  1. ikip
  2. |M
  3. okers
  4. |m
  5. тяж
  6. PageRoute
  7. oker
  8. |"
  9. ulfilled
  10. ��