Neuron Explorer

Qwen3-0.6B

blank slot before a number

A neuron in layer 4 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 spot right after a label like 'Number of pages:' or 'Word Count:' where a number should come next. It is a 'a figure goes here' signal. Turning it up did not produce a clear theme in the written 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

    from which they had been absent<|endoftext|>THE RAATZ ROUNDUP⏎Submitted for use on April 22, 2022⏎Word Count: 432⏎By State Sen. Jeff Raatz (R-Richmond

  2. #2 peak

    :59 PM MDT · Views: 1,130⏎HUMAN EMBRYONIC STEM CELL RE

  3. #3 peak

    <|endoftext|>The Art of Story Writing⏎by Nathaniel C. Fowler⏎Publisher: Sulley and Kleinteich 1913⏎Number of pages: 282⏎The writing of stories of every class and of any length,

  4. #4 peak

    vi, 80 p. :|⏎|Number of Pages||99|⏎|2||WCC exchange -- 2|⏎nodata File Size: 5MB.⏎Yet for many in the Roman Catholic church in Latin America, not

  5. #5 peak

    anna Climate).⏎The average temperature for the year in Mount Surprise is 73.0°F (22.8°C). The warmest month

  6. #6 peak

    Smith Colleges 2010⏎Number of pages: 137⏎These notes represent an attempt to develop a new computer graphics course

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 user is a bit of a problem. The user is a bit of a problem. The user is a bit of a problem. The user is a bit of a problem. The user is a bit of a problem. The user is

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

    % of the⏎Denver is the state's largest city and was, in 2002, the 26th largest US city. Its estimated 2002 population was 560,415, but

  2. #2 peak

    ostellar cloud freely and the temperature remains constant. It is then obtained for Jeans' Mass, MJ . The fact that MJ, the minimum mass able to collapse, increases when the collapse proceeds produces the

  3. #3 peak

    years when they were under the protection of Lord Thirumal these group of peoples are popularly known by the name of Yadavars. The fact that they are named as Yadavars is due to the reason that Lord Krishna belonged

  4. #4 peak

    !<|endoftext|>It may still be difficult for some people to accept that an activity such as gambling, which is not a drug, can be addictive. But the idea that addiction is confined to drug addiction rests on an outdated concept of what addiction is.

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. L2:371 excites
  2. L3:178 weakly excites
  3. L2:520 weakly excites
  4. L3:534 weakly excites
  5. L1:205 weakly suppresses
  6. L3:2281 weakly excites
  7. L2:782 weakly suppresses
  8. L0:1142 weakly suppresses
  9. L3:50 weakly excites
  10. L3:3 weakly suppresses
  11. L1:394 weakly suppresses
  12. L2:940 weakly suppresses
  13. L1:2480 weakly suppresses
  14. L0:1911 weakly excites
  15. L2:449 weakly excites
  16. L2:434 weakly suppresses
  17. L2:2798 weakly suppresses
  18. L2:1249 weakly excites
  19. L1:755 weakly suppresses
  20. L1:8 weakly excites
  21. L3:748 weakly suppresses
  22. L1:1193 weakly excites
  23. L0:932 weakly excites
  24. L0:997 weakly suppresses

Feeds later layers

  1. L10:1829 excites
  2. L9:291 excites
  3. L7:806 weakly suppresses
  4. L5:134 weakly excites
  5. L5:2027 weakly excites
  6. L13:1355 weakly suppresses
  7. L5:1544 weakly excites
  8. L11:681 weakly excites
  9. L21:488 weakly suppresses
  10. L10:205 weakly suppresses
  11. L5:1479 weakly suppresses
  12. L7:347 weakly excites unclear: mid-sentence function words
  13. L11:2321 weakly suppresses
  14. L15:175 weakly excites
  15. L16:2741 weakly suppresses
  16. L10:206 weakly excites colon or equals sign before a definition
  17. L19:206 weakly excites
  18. L10:463 weakly excites
  19. L24:2340 weakly suppresses
  20. L12:1008 weakly excites
  21. L9:1907 weakly excites
  22. L18:759 weakly suppresses
  23. L17:1244 weakly suppresses
  24. L11:502 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 -0.77 and -0.683 values between -0.68 and -0.583 values between -0.58 and -0.4910 values between -0.49 and -0.3934 values between -0.39 and -0.30103 values between -0.30 and -0.20331 values between -0.20 and -0.11918 values between -0.11 and -0.011492 values between -0.01 and 0.08629 values between 0.08 and 0.18218 values between 0.18 and 0.2782 values between 0.27 and 0.3626 values between 0.36 and 0.4612 values between 0.46 and 0.5511 values between 0.55 and 0.657 values between 0.65 and 0.743 values between 0.74 and 0.844 values between 0.84 and 0.932 values between 0.93 and 1.036 values between 1.03 and 1.125 values between 1.12 and 1.222 values between 1.41 and 1.501 values between 1.50 and 1.592 values between 1.59 and 1.691 values between 2.92 and 3.01offnegativepositiveoften

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. 多样
  2. azines
  3. どの
  4. めて
  5. どう
  6. いくら
  7. 本书
  8. 是什么呢
  9. 位
  10. nowhere

Suppresses

  1. ')]
  2. '></
  3. '>$
  4. '>";⏎
  5. '>
  6. ')</
  7. amped
  8. lider
  9. ző
  10. '>"