Neuron Explorer

Qwen3-0.6B

zero before a decimal point

A neuron in layer 25 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 digit 0 when it sits just before a decimal point, most often in statistics such as p < 0.001. Turning it up made the model output strings of symbols and equals signs, so that test gives little insight.

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

    3, 95% CI = 1.110-1.529), local citizens (p < 0.001) and officials/clerks (p = 0.

  2. #2 peak

    6%) compared to those of white-nosed coati (9.05%), being this difference significant in summer 2012 (P < 0.00001), summer (P =

  3. #3 peak

    the results were printed to three decimal places. When typing in the new conditions, Lorenz entered the rounded-off numbers (0.506 rather than 0.506127, a difference of less than 0.

  4. #4 peak

    to directly correlate the LLCCs length to the diameter of the inner tubes and determine the optimal diameter for the high yield growth of carbyne (about 0.71 nm). Our approach of HV high temperature annealing has a great

  5. #5 peak

    where c = (10/9)2 ≈ 1.2 and x = (9/10)log102 ≈ 0.97. The expected number of powers of 2 without zeroes starting from

  6. #6 peak

    22.91% of sampled diarrhea patients. The seasonal distribution of norovirus infections was different from non-norovirus patients (p < 0.001), with a half-year peak. Higher proportions of males (

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 idea that a single individual's actual effort to achieve a certain intent or intent under actual economic conditions, such as actual economic efforts, actual efforts to achieve a certain intent, actual efforts to achieve a certain intent, actual efforts to achieve 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

    working contact. The U.S. has operated “interest sections” in places like Cuba and Iran.⏎South Korea set up a liaison office in the North Korean city of Kaesong last year after several months of increased discussions and contacts with the North

  2. #2 peak

    leading a flight of fighters of the 39th Fighter Interceptor Squadron, 51st Fighter Interceptor Wing, based at K-13, near Suwon, Republic of South Korea. He was flying his usual airplane, North

  3. #3 peak

    Kentucky, two border states that officially remained in the Union, while the western counties of Virginia rejoined the North. The Confederacy's capital was moved from Montgomery to Richmond, Virginia, in May 1861, and regular presidential and

  4. #4 peak

    photo of Trump and Kim Jong Un on its front page.⏎Trump and Kim were asked by reporters about the possibility of opening a liaison office during their meeting last month in Hanoi, Vietnam. “It’s actually not a bad idea,” Trump said

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. L21:351 suppresses
  2. L20:373 excites
  3. L20:2980 suppresses
  4. L23:1291 excites semicolons and list-like citations
  5. L22:553 weakly suppresses
  6. L23:209 weakly suppresses
  7. L22:544 weakly excites
  8. L6:2561 weakly suppresses
  9. L3:236 weakly excites
  10. L21:162 weakly excites
  11. L18:1132 weakly excites
  12. L24:127 weakly excites
  13. L0:2583 weakly suppresses
  14. L18:335 weakly suppresses
  15. L21:1829 weakly excites
  16. L20:1391 weakly excites
  17. L9:1522 weakly suppresses
  18. L17:1134 weakly excites
  19. L20:206 weakly suppresses
  20. L15:2169 weakly excites
  21. L18:1187 weakly suppresses
  22. L18:775 weakly suppresses
  23. L8:692 weakly excites
  24. L10:32 weakly suppresses

Feeds later layers

  1. L27:682 excites
  2. L27:858 excites
  3. L26:582 excites
  4. L26:122 excites
  5. L27:2125 excites
  6. L27:646 excites
  7. L27:390 weakly excites
  8. L27:2906 weakly excites the letter r in the middle of words
  9. L27:753 weakly excites
  10. L26:2614 weakly excites
  11. L26:606 weakly excites
  12. L26:712 weakly suppresses
  13. L27:2214 weakly excites
  14. L26:2661 weakly excites
  15. L27:760 weakly excites
  16. L27:1603 weakly excites
  17. L27:1424 weakly excites
  18. L27:3022 weakly excites
  19. L26:1353 weakly suppresses
  20. L26:227 weakly excites
  21. L27:2808 weakly suppresses
  22. L27:2851 weakly excites
  23. L27:415 weakly excites
  24. L26:2455 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 -14.27 and -13.201 values between -13.20 and -12.141 values between -7.87 and -6.811 values between -6.81 and -5.745 values between -5.74 and -4.683 values between -4.68 and -3.612 values between -3.61 and -2.5514 values between -2.55 and -1.48414 values between -1.48 and -0.423349 values between -0.42 and 0.6598 values between 0.65 and 1.717 values between 1.71 and 2.783 values between 2.78 and 3.842 values between 3.84 and 4.912 values between 5.97 and 7.041 values between 9.17 and 10.241 values between 13.43 and 14.501 values between 17.69 and 18.761 values between 18.76 and 19.821 values between 27.28 and 28.34offnegativepositiveoften

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. =r
  3. =c
  4. =q
  5. =z
  6. =v
  7. =d
  8. :
  9. =
  10. =m

Suppresses

  1. .GetEnumerator
  2. Sous
  3. economic
  4. \application
  5. imizer
  6. hai
  7. 着他
  8. 真情
  9. her
  10. activex