Neuron Explorer

Qwen3-0.6B

the digit 1 in age ranges

A neuron in layer 25 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 responds to the number 1 when it starts an age or a range of ages, such as "ages 18-24" or "aged 18 and older". Its opposite side fires on the 1 in powers of ten like "10^8", and turning it down made the model write endless zeros.

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

    the two most dangerous activities your teen will ever participate in. Vehicle crashes are the number one cause of death for young adults ages 14-24!⏎Sign up for a free on-line parent program provided to

  2. #2 peak

    high enough to mean that 20% of all deaths in this age group in 2016 involved opioids.⏎Among those between the ages of 15 and 24, the report adds, the nearly 3,0

  3. #3 peak

    current study, healthy adults (age 18-35) read short discourse scenarios for comprehension while monitoring for the presence

  4. #4 peak

    in the late fall through winter. Some may experience SAD in the summer, but not as frequently. SAD is more common in young people (ages 18-30), particularly women. According to the Cleveland Clinic, higher risk

  5. #5 peak

    Cancer Society researchers led by Vilma Cokkinides, Ph.D., analyzed survey results from 4756 smokers (aged 18 and older) who visited a healthcare provider within the past year. All were

  6. #6 peak

    , year and sponsor across US cable networks and programs. Researchers calculated ad exposure for youth ages 12 to 17 years of age and young adults 18 to 24 years old.⏎The research showed that more than

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:000000000000000000000000000000000000000000000000

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

    = 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

  2. #2 peak

    substance of one density into a substance of different density, the speed of light changes and bending of light occurs.⏎In air, light travels at 3 x 108 m/sec but in water, it travels only about 2.2

  3. #3 peak

    water 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.

  4. #4 peak

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

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. L17:512 strongly suppresses
  2. L18:499 strongly excites
  3. L20:416 strongly excites the "1" in years 191x
  4. L3:244 suppresses
  5. L0:871 excites
  6. L4:247 suppresses
  7. L10:848 suppresses
  8. L19:720 excites
  9. L4:1401 excites unclear: letter fragments inside names
  10. L22:332 suppresses
  11. L21:518 excites
  12. L2:2029 suppresses
  13. L22:2733 suppresses
  14. L2:1866 excites
  15. L13:620 weakly excites
  16. L16:113 weakly suppresses
  17. L21:1031 weakly suppresses
  18. L5:559 weakly suppresses
  19. L20:287 weakly suppresses
  20. L10:722 weakly suppresses
  21. L21:271 weakly suppresses
  22. L2:1724 weakly excites
  23. L18:81 weakly suppresses
  24. L2:1699 weakly excites

Feeds later layers

  1. L26:413 suppresses
  2. L27:1153 suppresses
  3. L27:1196 suppresses
  4. L26:848 suppresses
  5. L27:153 excites the spot just before a historical number
  6. L27:205 suppresses
  7. L26:737 suppresses
  8. L27:862 suppresses
  9. L26:121 suppresses
  10. L27:396 suppresses
  11. L26:233 excites
  12. L27:238 suppresses
  13. L26:1461 suppresses
  14. L27:530 weakly excites
  15. L26:216 weakly suppresses blank slot before a year or number
  16. L27:1329 weakly excites
  17. L27:241 weakly excites
  18. L26:1065 weakly excites
  19. L27:1087 weakly excites
  20. L27:132 weakly suppresses
  21. L27:2659 weakly excites
  22. L26:598 weakly suppresses spot just before a year or century
  23. L26:905 weakly excites
  24. L27:902 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 -44.38 and -42.621 values between -39.12 and -37.361 values between -25.10 and -23.341 values between -16.33 and -14.581 values between -14.58 and -12.833 values between -12.83 and -11.071 values between -11.07 and -9.325 values between -9.32 and -7.575 values between -7.57 and -5.823 values between -5.82 and -4.0619 values between -4.06 and -2.3165 values between -2.31 and -0.563757 values between -0.56 and 1.2024 values between 1.20 and 2.959 values between 2.95 and 4.706 values between 4.70 and 6.452 values between 6.45 and 8.212 values between 8.21 and 9.961 values between 22.23 and 23.981 values between 23.98 and 25.73offnegativepositiveoften

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. HIP
  2. ��
  3. ourney
  4. logg
  5. developers
  6. 力
  7. เต
  8. tees
  9. Dere
  10. rack

Suppresses

  1. 0
  2. 0
  3. ₀
  4. ǒ
  5. 零
  6. OUSE
  7. од
  8. ouse
  9. če
  10. ор