Neuron Explorer

Qwen3-0.6B

zero in powers of ten

A neuron in layer 10 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 lights up on the zero in scientific notation, like the 0 in 3 x 10^8 or 10^67. When it was turned up, the model's writing drifted toward powers of ten and big-number maths.

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

    than in air, water is said to have a greater optical density than air.⏎Substance Speed of light (m/sec)⏎Vacuum 3 x 108⏎Air 3 x 108⏎Water 2.2

  2. #2 peak

    when written as 52!, but it can also be written in scientific notation as 8.065817517 × 10^67. If I attempt to write it using normal numerals, it

  3. #3 peak

    0, and mℓ = 0 is where σ = r/a0 and a0 is the Bohr radius (5.29 × 10-11 m). Calculate the position of the nodes for this wave function. Write

  4. #4 peak

    In air, light travels at 3 x 108 m/sec but in water, it travels only about 2.258 x 108 m

  5. #5 peak

    oltzmann formula:⏎E = εσT4⏎where ε=emissivity, a material property; σ = 5.67×10-8 ; T = temperature in Kelvin (absolute

  6. #6 peak

    6 inches. This radium emits tiny amounts of radon into the atmosphere. On average, one part of radon is present in 1 x 1021 part of air.Radon is also present in some spring waters, including

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 have been working on this. I have been working on it. I am ready to share it. I am ready to share this. I am ready to share this. Now, I will share it. I am ready to share this

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

    DLR, and NASA/JPL.⏎More information about Dawn is online at http://dawn.jpl.nasa.gov.⏎Image credit: NASA/JPL-Caltech/UCLA/MPS/DLR/IDA<|endoftext|>How To Guide⏎How to do a

  2. #2 peak

    our Big Blue Oasis in space:⏎Click on the globe to see enlargements of the 1400 or so photographs NASA received.⏎Courtesy NASA/JPL/University of ArizonaAn National Geographic website article says the next rover mission to the planet

  3. #3 peak

    voting rights, unless such rights are specifically withheld, as in special classes of “nonvoting” shares. The common stockholders’ legal rights may also include preemptive rights to maintain

  4. #4 peak

    93nm, deep in the ultraviolet range, and have been instrumental in driving the semiconductor industry from 90nm geometries down to 28nm. Unfortunately, 193

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. L9:1525 suppresses
  2. L9:2397 suppresses
  3. L8:2324 suppresses
  4. L9:5 excites
  5. L5:883 weakly excites
  6. L9:1740 weakly suppresses
  7. L9:2628 weakly excites
  8. L6:631 weakly excites
  9. L5:1608 weakly suppresses
  10. L8:2846 weakly excites
  11. L8:360 weakly suppresses
  12. L6:2689 weakly excites
  13. L3:1529 weakly suppresses
  14. L8:7 weakly excites
  15. L6:1112 weakly excites
  16. L8:16 weakly suppresses
  17. L7:1871 weakly excites
  18. L9:1916 weakly suppresses
  19. L0:1082 weakly suppresses
  20. L9:2976 weakly suppresses
  21. L1:1546 weakly suppresses
  22. L8:1533 weakly excites
  23. L8:1648 weakly suppresses
  24. L8:3 weakly excites

Feeds later layers

  1. L14:3032 excites
  2. L19:1242 excites
  3. L11:2380 excites
  4. L13:2285 excites
  5. L19:1542 excites
  6. L16:1248 weakly excites
  7. L23:2436 weakly excites
  8. L20:1668 weakly excites
  9. L19:2313 weakly excites
  10. L14:1875 weakly excites
  11. L23:2146 weakly suppresses
  12. L17:2479 weakly excites
  13. L12:2172 weakly excites
  14. L13:2915 weakly excites
  15. L18:2199 weakly suppresses
  16. L18:798 weakly excites
  17. L11:1106 weakly excites
  18. L11:2532 weakly excites
  19. L19:1365 weakly suppresses
  20. L21:2187 weakly excites
  21. L16:1325 weakly excites
  22. L25:895 weakly excites
  23. L16:1077 weakly excites
  24. L18:1029 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 -1.69 and -1.571 values between -1.57 and -1.451 values between -0.84 and -0.721 values between -0.72 and -0.597 values between -0.59 and -0.4739 values between -0.47 and -0.35107 values between -0.35 and -0.23392 values between -0.23 and -0.111300 values between -0.11 and 0.021368 values between 0.02 and 0.14539 values between 0.14 and 0.26109 values between 0.26 and 0.3829 values between 0.38 and 0.503 values between 0.50 and 0.632 values between 0.63 and 0.753 values between 0.75 and 0.871 values between 0.99 and 1.111 values between 1.24 and 1.361 values between 1.85 and 1.971 values between 2.09 and 2.211 values between 2.45 and 2.581 values between 3.06 and 3.19offnegativepositiveoften

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. кос
  2. 很久
  3. 亿
  4. phosphate
  5. ione
  6. osaur
  7. Jahres
  8. erals
  9. Podesta
  10. ศา

Suppresses

  1. sắc
  2. دن
  3. ordon
  4. chatt
  5. 和地区
  6. onStop
  7. ="%
  8. roat
  9. ADDING
  10. 内容