Neuron Explorer

Qwen3-0.6B

power and energy unit symbols

A neuron in layer 25 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 lights up on the part of a measurement that comes right before a unit, like the 'k' in kilowatts, the 'M' in MWh, or the 'W' in watts. When it was turned up, the model's replies drifted toward electricity and power talk, with a stray 'W' appearing, though the effect was otherwise messy.

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

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

  2. #2 peak

    therefore expressed as a quantity of energy per unit of time. The international unit for the power is in Joules/second (J/s) which is also known as Watts (W).⏎In electricity, the power is determined by the product of the

  3. #3 peak

    NNL's solid oxide fuel cell development program. "However, this research shows that smaller solid oxide fuel cells that generate between 1 and 100 kilowatts of power are a viable

  4. #4 peak

    California — home to about half of the EVs in the United States. They find that the arrival of an EV increases household electricity consumption by 2.9 kilowatt hours per day — less than half the amount assumed by state regulators.⏎

  5. #5 peak

    s, flush residues after fossil fuels are burned and cool power plants.⏎Coal uses up to 3.2 cubic metres of water per megawatt hour (MWh) of electricity produced, gas uses up to 1.7 m3

  6. #6 peak

    out a stage in the process and making the whole system more efficient because the energy is converted only once.⏎The group estimates that the concept could reduce the cost per MWh of offshore wind energy by a factor of four, avoiding added system costs associated

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: the big picture, the big picture, the big picture, the big picture... This is the big piece of the puzzle, right? The big piece of the puzzle, right? The big piece of the puzzle, right? The big piece

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

    ney. He wakes up the next day, only to discover that it's not the next day at all. It's Groundhog Day all over again.⏎via Wikimedia Commons⏎For some reason he's trapped in Feb. 2, forced to

  2. #2 peak

    remains of the dismantled schooner Wawona.⏎Built by influential shipbuilder Hans Ditlev Bendixsen, the 165 ft long schooner Wawona sailed from 1897 to 1947

  3. #3 peak

    connection. Here are the steps to create an ad hoc network in Windows 7.⏎- Click on Start (Windows icon) and type wireless. Click on Manage wireless networks.⏎- Click on Add to add a network.⏎- Click on Create an

  4. #4 peak

    emissions from aircraft engines. This means it may be worthwhile for planes to fly at lower altitudes, even if they do release more greenhouse emissions that way. Robert Noland and his team calculate that if planes flew low enough to

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. L24:1056 suppresses
  2. L23:895 excites
  3. L24:1520 weakly excites
  4. L8:3013 weakly excites
  5. L21:1978 weakly excites
  6. L13:1559 weakly suppresses
  7. L24:2592 weakly suppresses
  8. L20:2885 weakly suppresses
  9. L8:2206 weakly suppresses
  10. L4:1339 weakly suppresses
  11. L17:2350 weakly excites
  12. L22:1184 weakly suppresses
  13. L11:2779 weakly suppresses
  14. L7:2077 weakly suppresses
  15. L23:175 weakly suppresses
  16. L4:1652 weakly excites
  17. L20:31 weakly excites
  18. L0:1725 weakly excites
  19. L1:449 weakly excites
  20. L14:2744 weakly excites
  21. L23:1861 weakly excites
  22. L22:1701 weakly excites
  23. L13:2162 weakly excites
  24. L20:1392 weakly excites

Feeds later layers

  1. L27:848 strongly excites
  2. L26:556 strongly excites
  3. L27:457 strongly excites
  4. L27:670 excites
  5. L27:621 weakly excites
  6. L27:1625 weakly excites
  7. L26:973 weakly excites
  8. L26:1803 weakly suppresses
  9. L27:1121 weakly excites
  10. L27:2129 weakly suppresses
  11. L27:2707 weakly excites
  12. L26:3040 weakly excites
  13. L27:1013 weakly suppresses
  14. L27:2491 weakly suppresses
  15. L26:2677 weakly excites
  16. L27:1994 weakly suppresses
  17. L26:2967 weakly excites
  18. L27:2312 weakly suppresses
  19. L26:1167 weakly excites
  20. L26:1721 weakly suppresses
  21. L26:419 weakly excites
  22. L26:1960 weakly suppresses
  23. L26:2908 weakly suppresses
  24. L27:1593 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.

2 values between -18.05 and -17.251 values between -13.28 and -12.491 values between -12.49 and -11.691 values between -11.69 and -10.901 values between -10.90 and -10.101 values between -10.10 and -9.313 values between -9.31 and -8.522 values between -8.52 and -7.722 values between -7.72 and -6.934 values between -6.93 and -6.136 values between -6.13 and -5.345 values between -5.34 and -4.548 values between -4.54 and -3.7526 values between -3.75 and -2.9531 values between -2.95 and -2.1665 values between -2.16 and -1.37177 values between -1.37 and -0.572503 values between -0.57 and 0.22946 values between 0.22 and 1.0259 values between 1.02 and 1.8125 values between 1.81 and 2.6111 values between 2.61 and 3.406 values between 3.40 and 4.208 values between 4.20 and 4.992 values between 4.99 and 5.782 values between 5.78 and 6.581 values between 6.58 and 7.374 values between 7.37 and 8.172 values between 8.17 and 8.962 values between 10.55 and 11.341 values between 12.93 and 13.73offnegativepositiveoften

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. W
  2. W
  3. w
  4. w
  5. .w
  6. -w
  7. _w
  8. .W
  9. (W
  10. W

Suppresses

  1. stin
  2. essor
  3. utherford
  4. libs
  5. teenth
  6. отз
  7. 琯
  8. олж
  9. 第三个
  10. axe