Neuron Explorer

Qwen3-0.6B

mid-clause words in technical prose

A neuron in layer 17 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 fires on a word in the middle of a sentence that is about to be completed by the rest of the clause, such as 'means', 'changed', 'has' or 'that'. Its strongest moments come in plain explanations of how something works. Turning it up or down gave rambling or number-heavy text, not a clear theme.

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

    or mod)⏎operator to find the remainder one gets when two numbers are divided,⏎say like a % b. If a remainder is ever zero, it means that b⏎divides evenly into a, or a is divisible by b.⏎

  2. #2 peak

    hash of that data; on the receiving end, we re-hash the received data and check that the computed hash equals that sent; if any of the data has changed then (with overwhelming probability), the computed hash value will no longer match the original

  3. #3 peak

    server to the browser.⏎Like most other servers, Apache multiplexes its operations so that it can⏎start processing a new request before it has finished working on the previous⏎one. On Unix systems, Apache uses a multiprocess model in which

  4. #4 peak

    data file.⏎In the sub-matrix format a portion (the block size) of a row is followed by further portions of subsequent rows before the remaining data values of the row are found.⏎Of the two, the sub-matrix format provides for

  5. #5 peak

    completing a circuit.⏎This process continues until you turn off the light bulb.⏎What is Contact Current?⏎Contact current is a phenomenon that occurs when you touch something that has current flowing on it. As soon as you touch the object, the current goes

  6. #6 peak

    hormone insulin then facilitates the uptake of glucose into cells, where it is metabolized into energy for immediate use.⏎To conserve fuel, you body stores excess glucose not needed for energy as a compound called glycogen. Through a process

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 will start a new project, and I will write a project report, then I will write a project report, then I will write a project report, and then I will write a project report, and then I will write a project report

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

    When will the redistricting process begin?⏎The Census Bureau was delayed in its decennial count because of the COVID-19 pandemic. Typically, the bureau delivers data to the states by April 1.⏎This year, the bureau warned states

  2. #2 peak

    extensive research program as the industry addresses impacts and works toward 2025 industry goals⏎Every five years, Field to Market: The Alliance for Sustainable Agriculture publishes its National Indicators Report (NIR). The latest

  3. #3 peak

    image representing the strength and solidarity of Londoners emerged on Twitter. It’s not uncommon for workers on the London Underground to write messages of national unity on tube signs following tragic events, and the March terrorist attack seemed no different. The image proceeded to

  4. #4 peak

    Science article published this month, Americans should expect a tough flu season -- and an early one. While the article admits that scientists and doctors struggle to predict flu seasons each year, general consensus indicates that the 2017 - 2

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. L12:1702 excites
  2. L16:2363 excites
  3. L15:1829 excites
  4. L15:1113 suppresses
  5. L15:560 excites
  6. L14:2104 suppresses
  7. L14:1347 suppresses
  8. L13:708 suppresses
  9. L16:2375 suppresses
  10. L16:2332 suppresses
  11. L15:1641 excites
  12. L15:2724 suppresses
  13. L15:720 suppresses
  14. L14:2571 excites
  15. L13:563 excites
  16. L14:797 suppresses
  17. L12:937 excites
  18. L14:1520 excites
  19. L15:1879 excites
  20. L13:1153 excites
  21. L11:311 suppresses
  22. L16:1748 suppresses
  23. L13:2527 suppresses
  24. L15:689 suppresses

Feeds later layers

  1. L19:553 strongly excites
  2. L19:2563 strongly excites
  3. L19:1862 strongly suppresses
  4. L19:894 suppresses
  5. L18:2594 excites
  6. L19:2475 excites
  7. L21:1093 excites
  8. L24:60 suppresses
  9. L20:2968 suppresses 'more' or 'not' before a likelihood claim
  10. L22:2399 excites
  11. L18:1319 suppresses
  12. L22:1616 excites
  13. L19:2254 excites
  14. L18:3047 excites
  15. L24:1770 excites
  16. L19:2536 excites
  17. L24:778 excites
  18. L18:1005 excites
  19. L24:252 excites
  20. L26:886 excites
  21. L22:745 suppresses
  22. L19:2133 suppresses
  23. L21:958 excites
  24. L18:535 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 -9.42 and -9.091 values between -8.42 and -8.081 values between -7.75 and -7.412 values between -7.41 and -7.082 values between -6.74 and -6.411 values between -6.41 and -6.072 values between -6.07 and -5.744 values between -5.40 and -5.063 values between -5.06 and -4.732 values between -4.73 and -4.394 values between -4.39 and -4.068 values between -4.06 and -3.726 values between -3.72 and -3.3912 values between -3.39 and -3.0513 values between -3.05 and -2.7225 values between -2.72 and -2.3839 values between -2.38 and -2.0546 values between -2.05 and -1.7168 values between -1.71 and -1.38131 values between -1.38 and -1.04177 values between -1.04 and -0.71283 values between -0.71 and -0.37622 values between -0.37 and -0.041212 values between -0.04 and 0.30912 values between 0.30 and 0.63285 values between 0.63 and 0.9733 values between 0.97 and 1.312 values between 1.31 and 1.642 values between 1.64 and 1.982 values between 1.98 and 2.314 values between 2.31 and 2.651 values between 2.65 and 2.982 values between 3.65 and 3.99offnegativepositiveoften

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. 现在的
  2. 現在
  3. 如今
  4. now
  5. 而现在
  6. -now
  7. 现在
  8. 今
  9. 算了
  10. (now

Suppresses

  1. usually
  2. often
  3. sometimes
  4. 必ず
  5. 注明出处
  6. seldom
  7. usually
  8. yapılır
  9. 时常
  10. lops