Neuron Explorer

Qwen3-0.6B

comma after a full clause

A neuron in layer 19 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 a comma that comes right after a complete statement, where a new clause is about to start, like "struck by a distracted driver," or "emit a great amount of radiation,". When turned down, the model got stuck repeating "if" and "when".

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

    July, 2007. Lynn Grant and her daughters Shannon and Hannah were on the way to a birthday party when they were struck by a distracted driver, who ran a red light driving 55 mph. Hannah 6 years old

  2. #2 peak

    Thermostats regulate the system’s temperature, while filters clean the circulated air. Understanding these parts is crucial for effective maintenance and

  3. #3 peak

    As transistors, the building blocks of microprocessors, approach their minimum size limits, creative ways to continue increasing computing power have emerged. These include technologies such as pip

  4. #4 peak

    women starting at the age of 21, so she makes a mental note to ask specifically about any previous pap smears and enters the room to meet Stephanie, who is

  5. #5 peak

    It is the number of protons in the nucleus minus the number of electrons in between the nucleus and the electron in question. Basically, the nucleus attracts an electron, but other electrons in lower shells repel it (opposites attract, likes rep

  6. #6 peak

    diving real close to the side.<|endoftext|>Fashion and style are often used interchangeably, but they are two distinct concepts. While fashion refers to the trends and clothing items

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: by if to in when when when if if if if if if if if when when when when when if when when when when when if when when when when if when when when when if when when when when when if when when when when if

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

    . For years, researchers have sought to understand how such a small creature could make such a long journey, particularly because it entails crossing the Gulf of Mexico without stopping.⏎Stories told by generations past

  2. #2 peak

    ’re giving each student a chance to consider the question, weigh arguments for and against each answer choice, and commit to what s/he thinks is the best answer–all before s/he hears what the other students think about the question.⏎All students are

  3. #3 peak

    . Often this is the only test done which is just not sufficient, especially if you are experiencing symptoms of low thyroid function.⏎What it does: TSH

  4. #4 peak

    the laws of relativity.⏎For example, the Official Handbook of the Marvel Universe states that the character Northstar maintains speeds which can be considered "modest," especially when carrying a passenger. The Handbook also concedes that a solid object moving in

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. L17:3021 suppresses
  2. L14:35 suppresses
  3. L18:366 excites comma inside numbers and dates
  4. L18:93 suppresses
  5. L16:1059 excites
  6. L15:286 excites
  7. L16:846 excites comma opening an aside or explanation
  8. L16:797 suppresses
  9. L17:3068 excites comma after a sentence subject
  10. L4:3 excites commas after citation fields
  11. L16:240 weakly excites comma after 'sometimes' or 'the truth'
  12. L6:2362 weakly excites
  13. L18:2434 weakly suppresses
  14. L17:1597 weakly excites
  15. L11:154 weakly suppresses commas after citation numbers
  16. L5:16 weakly excites
  17. L17:1537 weakly excites
  18. L12:1295 weakly suppresses
  19. L17:585 weakly excites
  20. L11:2745 weakly suppresses
  21. L16:316 weakly suppresses
  22. L15:2835 weakly excites
  23. L15:630 weakly suppresses
  24. L17:5 weakly suppresses

Feeds later layers

  1. L24:148 strongly suppresses
  2. L20:458 strongly suppresses
  3. L21:140 excites
  4. L20:1245 suppresses
  5. L22:160 excites
  6. L21:1064 suppresses
  7. L23:449 suppresses
  8. L22:132 suppresses
  9. L24:116 excites
  10. L27:338 suppresses
  11. L23:756 suppresses
  12. L25:768 suppresses
  13. L22:478 suppresses
  14. L23:664 suppresses
  15. L26:852 suppresses
  16. L27:123 excites
  17. L25:626 suppresses
  18. L24:784 suppresses
  19. L27:511 suppresses
  20. L21:1597 suppresses
  21. L21:1023 suppresses
  22. L27:659 suppresses
  23. L26:509 suppresses
  24. L26:790 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.

2 values between -23.73 and -22.901 values between -17.09 and -16.262 values between -15.43 and -14.601 values between -13.77 and -12.941 values between -12.94 and -12.111 values between -12.11 and -11.272 values between -11.27 and -10.441 values between -8.78 and -7.954 values between -7.95 and -7.125 values between -7.12 and -6.294 values between -6.29 and -5.464 values between -5.46 and -4.639 values between -4.63 and -3.8011 values between -3.80 and -2.9722 values between -2.97 and -2.1435 values between -2.14 and -1.3193 values between -1.31 and -0.483290 values between -0.48 and 0.36353 values between 0.36 and 1.1919 values between 1.19 and 2.0218 values between 2.02 and 2.8511 values between 2.85 and 3.683 values between 3.68 and 4.516 values between 4.51 and 5.342 values between 5.34 and 6.172 values between 6.17 and 7.004 values between 7.00 and 7.831 values between 7.83 and 8.661 values between 8.66 and 9.49offnegativepositiveoften

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. ца
  2. 允许
  3. ,
  4. 拉
  5. 腾
  6. apest
  7. FAULT
  8. 半岛
  9. 是你
  10. ,t

Suppresses

  1. if
  2. when
  3. jika
  4. through
  5. via
  6. without
  7. during
  8. whenever
  9. when
  10. after