Neuron Explorer

Qwen3-0.6B

TV and video media words

A neuron in layer 23 of Qwen3-0.6B. Its strongest snippets are fairly consistent: several of them peak on the same token. It is featured, so it has steering outputs.

What this neuron does

This neuron fires on words about television, channels and video, such as 'TV', 'Channel 4' or the name of a broadcaster. When turned down, the model started writing about video and motion. Turned up, it wrote about unrelated topics.

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

    "Friends" (NBC), "Just Shoot Me" (NBC), "Ally McBeal" (FOX), "That '70s Show" (FOX), "The Simpsons" (FOX), and "Dawson's Creek

  2. #2 peak

    75% of e-cigarette ad exposure to youth occurred on cable networks, including AMC, Country Music Television, Comedy Central, WGN America, TV Land, and VH1. Researchers found that e-cigarette ads appeared on programs

  3. #3 peak

    marked the beginning of mainstream 3D for film and TV audiences – in fact only recently there were a spate of classic 3D movies played on Channel 4. These ‘glasses’ (the lenses in the cardboard design were a

  4. #4 peak

    exactly along the lines of history as we know it. Despite the book being banned, it has become an underground bestseller.⏎first season of Amazon’s highly rated TV series of the same name, roughly follows the outlines of Dick’s book with a

  5. #5 peak

    ysicist who writes for NPR’s 13.7 blog, described dark matter as being invisible (which makes it

  6. #6 peak

    it or discovered the alloy, and then used it to trap Warehouse agents Bering and Lattimer against the ceiling.⏎This is a good example of how the TV Universe is always in a state of flux, constantly being re-written. Until the

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: if the speed of sound can be recorded, then what would happen to the recorded motion of the moving object? How would it behave? What would it look like? How would it go? How would it move? What would it sound like

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

    player must support the codec used to encode the video data.⏎||MP4 files are an open standard that everyone can use.⏎|Programs that open AVI/MP4 files<|endoftext|>Pioneer Papers connects you with professional substantive editors to build and

  2. #2 peak

    model "VisMo", based on CIDOC CRM (University of Passau)<|endoftext|>||Audio Video Interleave File⏎||MPEG-4 Video File⏎||AVI is a video wrapper format created by Microsoft, storing video data that might

  3. #3 peak

    ‘Rigid Body’ and leave the type as ‘Active’. We want this to simulate animation. Now change the bounciness to 1.⏎- Playback the animation (Alt+A) and observe your bouncing ball! (Not related, but

  4. #4 peak

    , 2008, edition of The Boston Globe features an in-depth profile of Bisbee and his art. Click here to read the article and watch a

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. L20:1199 strongly suppresses
  2. L20:2621 excites
  3. L8:874 excites
  4. L14:700 excites
  5. L9:2779 excites
  6. L22:1292 excites
  7. L19:2326 suppresses
  8. L18:1766 suppresses
  9. L10:733 suppresses
  10. L4:2643 suppresses
  11. L8:2910 suppresses
  12. L6:2384 excites
  13. L22:2171 suppresses
  14. L7:1332 suppresses
  15. L7:2775 suppresses
  16. L9:2321 suppresses
  17. L17:1228 excites
  18. L22:2833 excites
  19. L21:1083 suppresses
  20. L9:617 suppresses
  21. L7:1822 suppresses
  22. L3:996 weakly suppresses
  23. L6:1914 weakly excites
  24. L20:2510 weakly excites

Feeds later layers

  1. L26:577 suppresses
  2. L26:2100 suppresses
  3. L26:2078 suppresses
  4. L24:2178 suppresses
  5. L27:2112 suppresses
  6. L24:2012 weakly suppresses
  7. L25:1683 weakly suppresses
  8. L26:269 weakly suppresses
  9. L24:2994 weakly suppresses
  10. L26:2547 weakly suppresses
  11. L26:966 weakly suppresses
  12. L25:1812 weakly suppresses
  13. L25:2044 weakly suppresses
  14. L27:2424 weakly suppresses
  15. L24:2722 weakly suppresses
  16. L26:2357 weakly suppresses
  17. L24:2325 weakly suppresses
  18. L27:308 weakly suppresses
  19. L25:1712 weakly suppresses
  20. L24:1657 weakly suppresses
  21. L24:2015 weakly suppresses
  22. L24:2732 weakly suppresses
  23. L24:2882 weakly suppresses
  24. L25:859 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 -24.53 and -22.771 values between -13.94 and -12.173 values between -12.17 and -10.414 values between -10.41 and -8.642 values between -8.64 and -6.888 values between -6.88 and -5.1120 values between -5.11 and -3.3483 values between -3.34 and -1.582724 values between -1.58 and 0.191030 values between 0.19 and 1.9524 values between 1.95 and 3.725 values between 3.72 and 5.481 values between 5.48 and 7.251 values between 7.25 and 9.021 values between 44.33 and 46.09offnegativepositiveoften

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. oyal
  2. AGMENT
  3. Sherlock
  4. Gree
  5. bourg
  6. фор
  7. TextAlign
  8. ќ
  9. ☒
  10. ISC

Suppresses

  1. video
  2. video
  3. 视频
  4. Video
  5. videos
  6. Video
  7. (video
  8. .Video
  9. _video
  10. videos