Neuron Explorer

Qwen3-0.6B

words of moderate or partial degree

A neuron in layer 8 of Qwen3-0.6B. Its strongest snippets are mixed: they peak on different tokens. It is featured, so it has steering outputs.

What this neuron does

This neuron responds to words that describe something in the middle or only sometimes, such as 'moderate', 'occasionally', 'partial', 'somewhat' and 'scattered'. Turning it up did not give writing that matched this theme in the test prompts.

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

    sorrel occurs in a freshwater marsh community and is common in annual grasslands, montane meadows and perennial bunchgrass communities.⏎Cal-IPC Inventory rating: Moderate⏎Cal-IPC Resources on Rumex acetosella⏎- California Invasive

  2. #2 peak

    Supervisor Districts 2 and 3.⏎Summers in the Calleguas Creek watershed are typically warm and dry, while winters are more mild with only occasional rainfall events

  3. #3 peak

    than the tubes and becoming reflexed.⏎Fruits: Berrylike, globose, 5-6 mm across, shiny red or purplish-black, occasionally brown, yellow, or whitish, but not glaucous, with

  4. #4 peak

    and new horizontally creeping shoots will soon form from the plant basis and gradually colonize the gravel.⏎|Light Demand||Low Light|⏎|Growth Rate||Moderate Growth|⏎Tissue culture plants should be removed from the cup as soon as possible

  5. #5 peak

    Back to 3⏎|5||Leaves mostly alternate or scattered; inflorescence a raceme of single flowers||Dendrophth

  6. #6 peak

    Engagement is highest at the elementary levels and declines fairly steadily through middle and high school, with a bit of a bump around 8th grade, as parents look towards high school.⏎3

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 just need to make sure that I get the best of the software, by using the best of the programming language, by using the best of the programming, I have no intention to make you understand the problem, so that you can be

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

    "...We Can Do It!". Or the the poster warning against spies with "...Loose Lips Sink Ships".⏎By the time you have completed our puzzle, you'll be, once again, very proud of the Greatest Generation.⏎Finished size 24

  2. #2 peak

    activity age appropriate and interesting).⏎From 2:30-3:30 PM we will be having fun with our "Structures & Materials" activity, where children and parents

  3. #3 peak

    wide must be present.⏎- Basement egress windows must also include window wells that provide enough clearance to allow escape.<|endoftext|>At the end of this lesson, students will be able to:⏎- Identify a geometric sequence.⏎- Graph a geometric sequence.⏎

  4. #4 peak

    of the language of mathematics as embodied in the basic constructs of mathematics in the fundamental areas of algebra, analysis, and geometry⏎- Be able to analyze the logical structure of a scientific or mathematical problem and to develop a meaningful

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. L6:1691 suppresses
  2. L7:1962 suppresses
  3. L7:870 excites
  4. L7:1170 weakly suppresses
  5. L7:1709 weakly excites
  6. L3:1672 weakly suppresses
  7. L2:2444 weakly excites
  8. L3:2697 weakly suppresses
  9. L6:1507 weakly suppresses
  10. L7:1922 weakly suppresses
  11. L1:311 weakly excites
  12. L4:1725 weakly excites
  13. L6:3002 weakly excites
  14. L6:1733 weakly excites
  15. L5:423 weakly suppresses
  16. L0:761 weakly excites
  17. L1:291 weakly excites
  18. L5:169 weakly excites
  19. L4:988 weakly suppresses
  20. L5:2497 weakly suppresses
  21. L4:2206 weakly suppresses
  22. L7:335 weakly excites
  23. L0:418 weakly suppresses
  24. L7:3015 weakly suppresses

Feeds later layers

  1. L18:3046 suppresses
  2. L19:627 weakly suppresses
  3. L17:2723 weakly suppresses
  4. L9:483 weakly suppresses
  5. L10:822 weakly excites
  6. L20:1857 weakly suppresses
  7. L9:2963 weakly excites
  8. L9:2518 weakly suppresses
  9. L13:520 weakly excites
  10. L9:426 weakly suppresses
  11. L9:292 weakly excites unclear: weak scattered tokens
  12. L12:427 weakly suppresses
  13. L25:1306 weakly suppresses comma after an intro phrase
  14. L24:551 weakly suppresses
  15. L10:1859 weakly suppresses
  16. L12:2271 weakly suppresses
  17. L19:650 weakly suppresses
  18. L9:2680 weakly suppresses
  19. L19:2739 weakly suppresses
  20. L25:2666 weakly suppresses
  21. L12:233 weakly excites
  22. L19:2508 weakly suppresses
  23. L13:1411 weakly excites
  24. L24:2149 weakly 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 -2.01 and -1.882 values between -1.19 and -1.061 values between -0.92 and -0.782 values between -0.78 and -0.657 values between -0.65 and -0.5114 values between -0.51 and -0.3763 values between -0.37 and -0.24353 values between -0.24 and -0.102285 values between -0.10 and 0.04889 values between 0.04 and 0.17193 values between 0.17 and 0.3142 values between 0.31 and 0.4521 values between 0.45 and 0.5812 values between 0.58 and 0.727 values between 0.72 and 0.865 values between 0.86 and 0.993 values between 0.99 and 1.131 values between 1.13 and 1.272 values between 1.40 and 1.542 values between 1.54 and 1.681 values between 1.95 and 2.091 values between 2.09 and 2.221 values between 3.32 and 3.45offnegativepositiveoften

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. 燧
  2. WithData
  3. 逗
  4. Await
  5. Decrypt
  6. ]/
  7. .with
  8. @
  9. with
  10. 见过

Suppresses

  1. isans
  2. 朋友们
  3. очных
  4. owitz
  5. admins
  6. ----------------------------------------------------------------------
  7. ants
  8. WXYZ
  9. RouteServiceProvider
  10. saida