Neuron Explorer

Qwen3-0.6B

'the' before a body part or organ

A neuron in layer 21 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 'the' when the next word is likely to name a part of the body, such as 'the lung', 'the liver', 'the neck' or 'the arteries'. Turning it up or down produced rambling text with no obvious body theme, so the steering results do not confirm the pattern.

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

    unborn fetus.⏎SEE ALSO: 10 Ways Mercury is Toxic to the<|endoftext|>Featured Photo: Crystalline caffeine. (Credit: Annie Cavan

  2. #2 peak

    to offer them.⏎There’s only scant data on how many patients who recover from Covid-19 are left with long-term fibrosis, or scarring of the lung. But studies on SARS and MERS, relatives of the novel coronavirus

  3. #3 peak

    have the option of joining the expedition for two weeks.⏎HOW WILL YOU HELP⏎You<|endoftext|>Adding and formatting mathematical equations in your papers can be a real pain in the neck. This is especially the case with complicated expressions involving matrices, calculus, bin

  4. #4 peak

    global temperatures rise and extreme heat waves become more common, so does heat illness. But while doctors have known for decades about the adverse effects of some medications on how the body regulates temperature, little is being done in the U.S. to better warn

  5. #5 peak

    appetite in the elderly.⏎- Hepatitis or chronic liver disease. Loss of appetite is known as one of the first symptoms associated with both hepatitis, or inflammation of the liver, and chronic liver disease.⏎- Depression. Psychological issues like depression,

  6. #6 peak

    life-threatening invasive cancer.⏎The researchers investigated a common type of early-stage breast cancer called ductal carcinoma in situ (DCIS). DCIS tumors are confined to the milk duct, and, though small, they are detectable by mammography.

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 two people the two people the two the two the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the 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

    shall be done with a soft but strong measuring tape just behind the shoulders. The person measuring shall always stand on the left side of the sheep, slide the end of the tape down the right side, pull it up the left side and pull the

  2. #2 peak

    to make your ball bounce down a path or trail really quick, just add a Wind force field (Shift+A -> Force Field -> Wind). Then set the strength of that force field to a point where you can see your ball move in the direction

  3. #3 peak

    raised by a pull of 100 lb on the free end of the rope. To raise the load 10 ft, however, the free end of the rope must be pulled 30 ft.⏎Disregarding friction, work

  4. #4 peak

    ave. A single sheave mounted in a block and fixed in place simply changes the direction of force exerted on the rope passing over it. If the end of the rope that ordinarily would attach to the load is passed around a second, unf

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. L20:47 excites
  2. L20:2520 excites
  3. L19:768 excites
  4. L20:837 suppresses
  5. L19:1361 excites
  6. L20:1494 excites
  7. L19:1817 excites
  8. L19:1181 excites
  9. L20:469 excites
  10. L18:1342 suppresses
  11. L19:656 suppresses
  12. L20:1972 suppresses 'of' in 'of course'
  13. L8:636 suppresses
  14. L19:2283 suppresses
  15. L19:1953 excites
  16. L20:219 excites "the" after time or place prepositions
  17. L19:2306 excites
  18. L19:685 excites
  19. L19:1056 excites
  20. L19:1785 weakly excites
  21. L20:762 weakly excites
  22. L17:429 weakly excites
  23. L18:1691 weakly suppresses
  24. L19:705 weakly suppresses

Feeds later layers

  1. L27:127 strongly excites
  2. L26:507 strongly suppresses
  3. L27:399 strongly suppresses
  4. L24:49 excites
  5. L23:94 excites
  6. L22:199 suppresses
  7. L25:2531 suppresses
  8. L22:171 weakly suppresses
  9. L27:994 weakly excites
  10. L27:121 weakly excites
  11. L26:251 weakly excites
  12. L25:1796 weakly suppresses
  13. L26:2278 weakly excites
  14. L25:2497 weakly suppresses
  15. L24:2538 weakly excites
  16. L27:2537 weakly excites
  17. L27:1464 weakly excites
  18. L27:1382 weakly suppresses
  19. L24:1050 weakly suppresses
  20. L23:1011 weakly excites
  21. L25:530 weakly suppresses
  22. L22:1589 weakly suppresses
  23. L24:837 weakly suppresses
  24. L24:1366 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 -10.16 and -9.752 values between -9.75 and -9.341 values between -8.52 and -8.121 values between -8.12 and -7.712 values between -7.71 and -7.301 values between -6.89 and -6.482 values between -6.48 and -6.072 values between -6.07 and -5.674 values between -5.67 and -5.265 values between -5.26 and -4.855 values between -4.44 and -4.037 values between -4.03 and -3.626 values between -3.62 and -3.2213 values between -3.22 and -2.8117 values between -2.81 and -2.4025 values between -2.40 and -1.9936 values between -1.99 and -1.5854 values between -1.58 and -1.1791 values between -1.17 and -0.77315 values between -0.77 and -0.362100 values between -0.36 and 0.05883 values between 0.05 and 0.46151 values between 0.46 and 0.8769 values between 0.87 and 1.2840 values between 1.28 and 1.6825 values between 1.68 and 2.0910 values between 2.09 and 2.5012 values between 2.50 and 2.9110 values between 2.91 and 3.324 values between 3.32 and 3.735 values between 3.73 and 4.133 values between 4.54 and 4.952 values between 4.95 and 5.361 values between 5.36 and 5.773 values between 5.77 and 6.18offnegativepositiveoften

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. zich
  3. 己
  4. {_
  5. thereof
  6. .jetbrains
  7. -neck
  8. HomeController
  9. 在家里
  10. 在那里

Suppresses

  1. the
  2. the
  3. The
  4. _the
  5. →the
  6. The
  7. .the
  8. THE
  9. -the
  10. ,the