Neuron Explorer

Qwen3-0.6B

unclear: weak hits on 'holiday' and 'playing', negative on 'Ada-fruit'

A neuron in layer 0 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 is weak and hard to pin down. It gives small boosts to words like 'holiday', 'playing' and 'indoor', and it is pushed strongly down by the 'fruit' in company names like Adafruit. Turning it up made the model repeat words, so it does not show a clear meaning.

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

    sponsored by the Cuban American National Foundation, Walesa started his speech by asking the crowd

  2. #2 peak

    Holiday Crowd At The Zoo⏎- The Child Is Father Of The Man (Es⏎

  3. #3 peak

    holiday. Sometimes the log would burn until the New Year.⏎During the days of rest

  4. #4 peak

    Playing a large role in the history of the early colonist, Native American tribes taught

  5. #5 peak

    flash. The sun is huge, but so far from the Earth that it acts as a hard light source. Hard light produces the best shadows for photos.⏎Outdoors, the lower the sun is in the sky, the longer the shadows objects here

  6. #6 peak

    indoor evaporator coil. The heat laden freon then flows back out to the outdoor

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:peciallypeciallypeciallypeciallypeciallypeciallypeciallypeciallypeciallypeciallypeciallypeciallypeciallypeciallypeciallypeciallypeciallypeciallypeciallypeciallypeciallypeciallypeciallypeciallypeciallypeciallypeciallypeciallypeciallypeciallypeciallypeciallypeciallypeciallypeciallypeciallypeciallypeciallypeciallypeciallypeciallypeciallypeciallypeciallypeciallypeciallypeciallypecially

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

    ayyah Painted Ware<|endoftext|>Firearms Transfer Stamps⏎On June 26, 1934, President Franklin D. Roosevelt signed the National Firearms Act of 1934 into law. The act placed a tax on

  2. #2 peak

    , “If we cannot celebrate a man who died for love, then how can we say we believe in it? It is up to me and you.”⏎Adafruit publishes a wide range of writing and video content, including interviews and reporting on the

  3. #3 peak

    and positive voice within the community – check it out here: adafruit.com/editorialstandards⏎Stop breadboarding and soldering – start making immediately! Adafruit’s Circuit Playground is jam-packed with LEDs, sensors, buttons, alligator clip

  4. #4 peak

    better understand the economic impact of North Carolina’s seafood industry,” says Jane Harrison, North Carolina Sea Grant’s coastal economics specialist and a project lead. “The shellfish results indicate the strength of the growing aquaculture sector.”⏎Supply and demand of

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

None: this is the first layer.

Feeds later layers

  1. L1:460 strongly excites
  2. L1:1262 strongly excites
  3. L1:2147 strongly excites
  4. L2:923 strongly excites second half of 'side by side' and 'face-to-face'
  5. L1:110 excites
  6. L1:81 suppresses
  7. L1:1321 excites
  8. L1:2583 excites
  9. L1:1662 excites
  10. L1:654 excites
  11. L1:1245 excites
  12. L1:1929 excites
  13. L1:1210 excites
  14. L1:615 excites
  15. L1:2677 excites
  16. L1:2099 excites
  17. L1:2159 excites
  18. L1:37 excites
  19. L1:2532 excites
  20. L1:1613 excites
  21. L1:1337 excites
  22. L1:2432 excites
  23. L1:65 excites
  24. L1:1498 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 -1.17 and -1.141 values between -0.39 and -0.352 values between -0.35 and -0.315 values between -0.28 and -0.2413 values between -0.24 and -0.2120 values between -0.21 and -0.1757 values between -0.17 and -0.1477 values between -0.14 and -0.10180 values between -0.10 and -0.06389 values between -0.06 and -0.031437 values between -0.03 and 0.011076 values between 0.01 and 0.04442 values between 0.04 and 0.08143 values between 0.08 and 0.1244 values between 0.12 and 0.1515 values between 0.15 and 0.195 values between 0.19 and 0.221 values between 0.22 and 0.26offnegativepositiveoften

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. ordinator
  2. sabotage
  3. @js
  4. valuator
  5. usterity
  6. ApiController
  7. pute
  8. onDelete
  9. 磺
  10. apur

Suppresses

  1. ara
  2. hea
  3. arda
  4. ío
  5. chair
  6. aba
  7. eload
  8. 庭
  9. ifes
  10. اح