Neuron Explorer

Qwen3-0.6B

'first' and 'inaugural' as starting points

A neuron in layer 11 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 words that mark a beginning, like 'your very first project', 'first birthday' or 'inaugural season'. When it was turned up, the model's writing often drifted toward talk about beginnings and firsts.

Written by Claude from this neuron's activations and steering outputs. Confidence: high. 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

    –75°F. Successive Sowings: Every 3 weeks until 10 to 12 weeks before your average first fall frost date for continuous harvest. Mild Climates: Sow in late fall

  2. #2 peak

    by the Fédération Internationale de l'Automobile (FIA). The FIA Formula One World Championship has been the premier form of racing since the inaugural season in 1950, although other Formula One races were regularly held

  3. #3 peak

    basics to help you create your very first project.⏎The first thing you want to do is make a new composition. The

  4. #4 peak

    the beginning, explaining the history of Java and why it's⏎important. You then learn how to obtain the Java Development Kit (JDK)⏎and write your first Java program.⏎Next, it's on to the Java fundamentals, including data types

  5. #5 peak

    , repeating sounds, and finger feeding.⏎- Step 2: Watch the child try to stand up with help from the furniture from just before the end of their first year, and finally walk with some help. They will bring objects to you,

  6. #6 peak

    purees with a spoon, starting at around 6 months of age when they can sit up and grab things.⏎However, some high-risk babies may need to start earlier due to allergy risks, and premature or specially challenged babies may need to start

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: I need to be able to run a system, and the user has to be logged out of the app. So I need to be able to access the data, and the system is running, and the user is logged in, and 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

    also shares her innovative methodology to help measure brown fat deposits that aid in non-shivering thermogenesis.<|endoftext|>Gruoch of Scotland⏎Gruoch (* 1015) was the daughter of Bodhe, Prince of Scotland, and

  2. #2 peak

    87), but only after a dry period. After wet periods, the risk of diarrhea actually decreased with such heavy rainfall events (incidence rate ratio = 0.74,

  3. #3 peak

    return-to-sport time of 14 days or longer had significantly higher total tau levels at 24-48 hours post injury (mean difference –0.51 pg/mL, 95% confidence interval, –0.

  4. #4 peak

    to a stable location centered near the geographic north pole. The authors find that leading up to the magnetic reversal, the strength of the magnetic field weakened to roughly 10 - 20

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. L9:1895 suppresses
  2. L7:502 suppresses unclear: words after 'less than' or 'of'
  3. L9:563 excites
  4. L6:389 excites
  5. L4:867 suppresses
  6. L4:2554 suppresses
  7. L7:844 suppresses
  8. L6:2226 suppresses
  9. L9:2152 excites
  10. L9:1348 suppresses
  11. L7:382 suppresses
  12. L3:2239 excites
  13. L8:1388 suppresses
  14. L10:2452 suppresses
  15. L8:766 excites
  16. L10:1284 excites
  17. L10:912 excites
  18. L10:2733 excites
  19. L8:1907 excites
  20. L8:545 suppresses
  21. L9:1032 suppresses
  22. L7:998 suppresses
  23. L8:2829 excites
  24. L5:3013 excites

Feeds later layers

  1. L16:885 strongly excites
  2. L12:455 excites
  3. L15:1916 excites
  4. L13:2098 suppresses
  5. L13:386 suppresses
  6. L14:2632 suppresses
  7. L13:292 excites
  8. L17:2811 excites
  9. L18:2221 excites
  10. L16:2909 excites
  11. L12:1935 excites
  12. L16:1120 excites
  13. L14:1809 excites
  14. L13:2656 excites
  15. L16:1188 suppresses
  16. L21:1691 excites "at" in "at first" and "initial" words
  17. L16:911 suppresses
  18. L16:1060 suppresses
  19. L13:919 excites
  20. L12:1272 suppresses
  21. L18:2298 excites
  22. L13:791 excites
  23. L18:1899 excites
  24. L13:2314 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 -2.26 and -2.021 values between -1.53 and -1.292 values between -1.29 and -1.045 values between -1.04 and -0.808 values between -0.80 and -0.56213 values between -0.56 and -0.311926 values between -0.31 and -0.071490 values between -0.07 and 0.18125 values between 0.18 and 0.4253 values between 0.42 and 0.6617 values between 0.66 and 0.9114 values between 0.91 and 1.1515 values between 1.15 and 1.4012 values between 1.40 and 1.645 values between 1.64 and 1.882 values between 1.88 and 2.131 values between 2.13 and 2.373 values between 2.37 and 2.611 values between 2.61 and 2.862 values between 2.86 and 3.103 values between 3.10 and 3.351 values between 3.35 and 3.592 values between 3.59 and 3.831 values between 4.08 and 4.321 values between 4.81 and 5.052 values between 6.03 and 6.271 values between 6.51 and 6.761 values between 7.25 and 7.49offnegativepositiveoften

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. 初始
  2. 开端
  3. 最初的
  4. :start
  5. 朴素
  6. basics
  7. beginnings
  8. (initial
  9. kickoff
  10. /simple

Suppresses

  1. 더라
  2. dens
  3. fuller
  4. �新
  5. maf
  6. >X
  7. (tm
  8. uffer
  9. ย
  10. N