Neuron Explorer

Qwen3-0.6B

'unless' after a prohibition

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 responds to the word 'unless' when it comes right after a rule or prohibition, as in 'should not be used unless' or 'do not adjust your dose unless'. Its strongest negative responses are on parts of Chinese names like 'China' and 'Chiang'. Turning it up did not bring out a clear theme in the sample text.

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

    manager from this point to ensure the employee’s contributions are valued and that their needs in knowing how to do their job are<|endoftext|>No person shall hold any civil office unless that person be a qualified elector for such office. Article IX, Section

  2. #2 peak

    women who weigh more or who exercise regularly.⏎Very low calorie diets of fewer than 800 calories per day should not be used unless you are being monitored by your doctor.⏎Healthy Weight Tip⏎Set goals to eat

  3. #3 peak

    medical condition being treated⏎- Other medical conditions you may have⏎- Other medications you may currently be taking.⏎As with any medication, do not adjust your dose unless your healthcare provider specifically instructs you to do so.⏎(Click Monopril

  4. #4 peak

    recommended, this is the final policy that was adopted by our pack.⏎Pack Technology Policy⏎Cub Scouts should not use electronic devices at Cub Scout meetings and events unless the activity requires the use of them

  5. #5 peak

    the same principle as the coil in an electric heater. There is a black spot on the metal at the back of the stove. Water heater elements must never operate unless immersed in water. A sudden voltage increase to the element caused by a power surge

  6. #6 peak

    resident for the four years immediately preceding election or appointment. One must also be at least twenty-five years old upon assuming office. No person can serve as attorney general unless he or she has been an active member of the State Bar of Georgia for at

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 question is to find the answer in a certain way. The answer is to write the code as a program that reads the input, and then outputs the function. The first argument is the list of numbers, and the second argument is 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

    years, Chinese currency existed in both Ancient and Imperial China. In 1914, the Silver Dollar was established as the official currency of the Republic of China, with copper, fen, and nickel coins being added in the 19

  2. #2 peak

    ung 41349, Taiwan (corresponding author's e-mail: firstname.lastname@example.org, email@example.com).⏎Chia-Liang Chiang is with

  3. #3 peak

    e-mail: firstname.lastname@example.org).⏎Cite:Ming-Ju Lee, Ming-Gin Lee, Yishuo Huang, and Chia-Liang Chiang, "Purification Study of Pervious Concrete Pavbement," International

  4. #4 peak

    use synchrotrons – extremely powerful x-ray sources – to trace the complex and intricate structure of the brain.⏎Professor Yeu-Kuang Hwu from the Institute of Physics, Academia Sinica and Professor Ann-Shyn Ch

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. L4:2588 excites
  2. L3:1396 excites
  3. L5:64 weakly excites
  4. L2:1488 weakly excites
  5. L5:2650 weakly excites
  6. L5:2497 weakly suppresses
  7. L5:2343 weakly suppresses the word "hair"
  8. L3:2315 weakly suppresses
  9. L5:102 weakly excites
  10. L1:2983 weakly excites
  11. L3:2973 weakly excites
  12. L4:1209 weakly suppresses
  13. L5:1462 weakly excites
  14. L4:577 weakly suppresses
  15. L2:126 weakly excites
  16. L4:2852 weakly excites
  17. L2:1766 weakly excites
  18. L4:3014 weakly suppresses
  19. L5:1023 weakly suppresses
  20. L4:1932 weakly suppresses
  21. L2:2700 weakly suppresses
  22. L0:3006 weakly excites
  23. L3:14 weakly excites
  24. L2:2763 weakly suppresses

Feeds later layers

  1. L8:2581 suppresses
  2. L9:1462 suppresses
  3. L16:219 weakly suppresses
  4. L7:829 weakly excites
  5. L14:223 weakly suppresses
  6. L8:1147 weakly excites
  7. L12:2987 weakly excites
  8. L11:1184 weakly suppresses
  9. L26:691 weakly excites
  10. L8:2418 weakly excites
  11. L7:2145 weakly suppresses
  12. L16:2891 weakly suppresses
  13. L19:243 weakly suppresses
  14. L23:2650 weakly suppresses
  15. L24:65 weakly suppresses
  16. L12:249 weakly suppresses
  17. L20:111 weakly suppresses
  18. L10:879 weakly suppresses
  19. L23:1146 weakly excites
  20. L9:1331 weakly suppresses
  21. L13:135 weakly suppresses document boundary end-of-text token
  22. L14:2771 weakly excites
  23. L7:444 weakly suppresses
  24. L13:1448 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 -3.10 and -2.961 values between -2.83 and -2.691 values between -0.93 and -0.803 values between -0.66 and -0.5322 values between -0.53 and -0.39173 values between -0.39 and -0.26718 values between -0.26 and -0.121641 values between -0.12 and 0.01987 values between 0.01 and 0.15277 values between 0.15 and 0.2860 values between 0.28 and 0.4215 values between 0.42 and 0.564 values between 0.56 and 0.692 values between 0.69 and 0.831 values between 0.96 and 1.101 values between 1.50 and 1.641 values between 2.18 and 2.31offnegativepositiveoften

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. 艿
  2. esson
  3. inston
  4. byname
  5. &W
  6. mascul
  7. USE
  8. 对其
  9. teams
  10. helmet

Suppresses

  1. 代
  2. idend
  3. iversal
  4. ####⏎
  5. fers
  6. fections
  7. ected
  8. ],[
  9. itud
  10. cles