Neuron Explorer

Qwen3-0.6B

the total after "N of" in fractions

A neuron in layer 8 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

Fires on the blank spot right after "of" in phrases like "Page 1 of", "out of" or "10 of the", where the model is about to write the total number. The strongest firing in the set comes from these spots. Turning it up did not bring out a matching theme in the generated 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

    2 – August 20.<|endoftext|>Genetics as it applies to evolution, molecular biology, and medical aspects.⏎10 posts • Page 1 of 1⏎Hope somebody can help, I've got an A level biology exam on

  2. #2 peak

    of bone that are not subject to great mechanical stress. It makes up much of ... (100 of 310 words)<|endoftext|>How do crabs usually die?⏎How do cr

  3. #3 peak

    24 attendees – aged one to 56. Twenty-two of the victims needed hospitalization. As of May 2nd, 10 of the 22 were still in hospital and in stable condition.⏎Kalikot District,

  4. #4 peak

    analysis. The number of answer matches between any pair of exam papers can be compared with the expected number of answer matches. Example: John (62 correct out of 100) and Barbara (67 correct out of 10

  5. #5 peak

    be introduced. Generalized anxiety disorder is estimated to occur in 5% of the general population. Women are generally more affected than men .⏎Page 1 of 2⏎Dating Advice For Men<|endoftext|>Feb 5, 2020

  6. #6 peak

    025.SS UNSPSC Code 30260000 See more.Customers who bought this item also bought.Page 1 of 1 Start over Page 1 of 1 .Previous page.1/4

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 history of the development of the United States, the years 1965–1900, and the Truman and postwar years. This is the only basis for your analysis. What do you mean? — I can

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

    know about the dangers of fires at home. Many households take active steps toward protecting family members from the risks of fire. This includes installing smoke detectors, keeping fire extinguishers charge, and establishing an emergency plan in case family members need to exit

  2. #2 peak

    and the personal pursuit of happiness. Undue taxation that provides pork for a<|endoftext|>Find All Factors⏎A sure way to find all factors of a number using prime factorization.⏎Algebra.com [offsite link]⏎Algebra.com has a variety

  3. #3 peak

    SE) = 10 log10(MAX/(MSE)^(1/2))^2 = 20 log10(MAX/(MSE)^(1/2)). Therefore, PSNR = <|endoftext|>In my current position as

  4. #4 peak

    102 ≈ 0.97. The expected number of powers of 2 without zeroes starting from the power N is cxN/(1-x) ≈ 40 ⋅ 0.97N.⏎Let

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. L4:1432 excites
  2. L5:1006 excites
  3. L7:1476 weakly excites
  4. L5:110 weakly suppresses
  5. L6:419 weakly excites
  6. L2:2044 weakly excites
  7. L5:1212 weakly suppresses
  8. L1:652 weakly suppresses
  9. L2:2188 weakly suppresses
  10. L6:922 weakly suppresses
  11. L6:1046 weakly suppresses
  12. L1:440 weakly suppresses
  13. L7:1428 weakly suppresses
  14. L0:1423 weakly excites
  15. L2:2701 weakly excites
  16. L5:504 weakly suppresses
  17. L2:217 weakly excites
  18. L5:27 weakly excites
  19. L6:92 weakly excites
  20. L5:1159 weakly excites
  21. L1:437 weakly suppresses the digit 9 at a fragment start
  22. L1:2161 weakly excites
  23. L7:1982 weakly suppresses
  24. L0:55 weakly excites

Feeds later layers

  1. L9:2439 excites
  2. L21:2855 excites
  3. L10:1700 excites
  4. L22:1627 excites
  5. L11:2679 excites
  6. L15:2436 excites 'out of' in ratios and rankings
  7. L18:2519 excites
  8. L12:2889 weakly excites
  9. L15:2428 weakly excites
  10. L19:1851 weakly excites
  11. L12:196 weakly excites
  12. L17:2906 weakly excites
  13. L12:472 weakly excites
  14. L17:2984 weakly excites blank before a number in prevalence claims
  15. L9:679 weakly excites
  16. L12:3001 weakly excites
  17. L10:1329 weakly excites
  18. L18:1329 weakly suppresses
  19. L10:2016 weakly excites
  20. L12:2289 weakly excites 'of' after part-whole nouns
  21. L14:2674 weakly excites
  22. L15:3034 weakly excites
  23. L15:1926 weakly excites
  24. L10:1926 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.

3 values between -0.86 and -0.751 values between -0.75 and -0.643 values between -0.64 and -0.5311 values between -0.53 and -0.4283 values between -0.42 and -0.31281 values between -0.31 and -0.20780 values between -0.20 and -0.091472 values between -0.09 and 0.02900 values between 0.02 and 0.13272 values between 0.13 and 0.2464 values between 0.24 and 0.3523 values between 0.35 and 0.465 values between 0.46 and 0.573 values between 0.57 and 0.682 values between 0.68 and 0.791 values between 1.12 and 1.231 values between 1.34 and 1.451 values between 1.67 and 1.781 values between 2.77 and 2.881 values between 3.43 and 3.54offnegativepositiveoften

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. 总量
  2. setContent
  3. 主流
  4. 共
  5. issent
  6. 总数
  7. 톱
  8. 大全
  9. 占比
  10. 全程

Suppresses

  1. eday
  2. nger
  3. responseData
  4. schlä
  5. agged
  6. iode
  7. 颛
  8. awai
  9. ��
  10. ffiti