Neuron Explorer

Qwen3-0.6B

metric or imperial conversions in brackets

A neuron in layer 23 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 fires on the opening bracket that comes right after a measurement, such as "700 kilometers (435 miles)", where a conversion to the other unit system follows. Turning it up gave text about numbers and quantities, but it does not show a clean, unique theme.

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

    , 2011. This image was taken through the camera’s clear filter. The distance to the surface of Vesta is 700 kilometers (435 miles) and

  2. #2 peak

    and Jeff had the privilege of naming it, and the short version is he decided it was my time.⏎The asteroid is about 1.3 km (0.8 miles) across, making it rather

  3. #3 peak

    )|⏎|• Water||83 sq mi (210 km2) 13%%|⏎|• Density||401/sq mi (155/km2)|⏎|Time zone||UTC−7 (

  4. #4 peak

    moved off their launch pads to take shelter from hurricanes and tropical storms.⏎But the shuttle cannot roll back to the VAB in winds exceeding 74 kph (46 mph), so NASA would need to move it well ahead of the storm

  5. #5 peak

    Highest point||top of flag staff|⏎|- elevation||15.75 ft (4.8 m) over stone base|⏎July 3, 19

  6. #6 peak

    50 feet (16 meters) away from the spacecraft, and had just been separated about three seconds earlier.⏎The heat shield is 15 feet (4.5 m) wide

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 an an example of a real-world work of work. Which of the following items would be a good example of work of work? (Select all of work of work items, as necessary) A. a piece of real-world work

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

    Indian and Atlantic oceans. The city’s elevation ranges from 5,700 to 5,930 feet (1,740 to 1,810 metres).⏎Aside from a few small streams and artificial lakes

  2. #2 peak

    , and flows generally southwestward, watering the Paris Basin, to join the Seine River at Conflans, after a course of 188 miles (302 km). It traverses the Thiérache Hills, where it

  3. #3 peak

    of beautiful new forms not seen before in inhabited structures.⏎Our final image is of the Chesa Futura in St. Moritz, Switzerland, a futuristic ski chalet which was built using many of the same automated wood-shaping technologies as are

  4. #4 peak

    ⏎A typical bedroom in the rest house⏎Inside the toilet block⏎The total cost of the building construction was VND 388 Million (approximately AUD 320<|endoftext|>Go Math Worksheets – Are you aware how to optimize using numbers

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. L22:948 strongly suppresses
  2. L22:1774 suppresses
  3. L22:2054 suppresses
  4. L22:2430 excites
  5. L22:474 excites
  6. L22:3034 suppresses
  7. L21:2040 suppresses
  8. L8:1245 weakly excites
  9. L5:967 weakly suppresses
  10. L19:3029 weakly excites
  11. L5:2760 weakly excites
  12. L11:2505 weakly suppresses
  13. L19:2275 weakly suppresses
  14. L21:1601 weakly suppresses
  15. L22:462 weakly excites
  16. L7:838 weakly excites
  17. L3:978 weakly excites
  18. L19:1259 weakly excites
  19. L21:138 weakly suppresses
  20. L11:2260 weakly suppresses
  21. L0:1932 weakly excites
  22. L17:2227 weakly suppresses
  23. L16:1183 weakly excites
  24. L6:1382 weakly suppresses

Feeds later layers

  1. L24:76 suppresses
  2. L24:2765 suppresses
  3. L26:2679 weakly suppresses
  4. L24:2296 weakly excites
  5. L25:230 weakly excites
  6. L24:952 weakly excites
  7. L24:342 weakly excites unit conversions in parentheses
  8. L24:2194 weakly excites
  9. L24:2812 weakly suppresses
  10. L25:2029 weakly excites
  11. L24:408 weakly excites
  12. L26:739 weakly excites
  13. L27:391 weakly excites
  14. L24:2657 weakly suppresses
  15. L24:2527 weakly suppresses
  16. L25:2519 weakly excites
  17. L24:2854 weakly suppresses
  18. L25:1749 weakly suppresses
  19. L26:463 weakly suppresses digits inside years and numbers
  20. L24:2895 weakly excites
  21. L27:1044 weakly suppresses
  22. L24:2911 weakly excites the 'cul' fragment in culprit and culminate
  23. L26:1698 weakly suppresses
  24. L24:2481 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 -15.06 and -14.271 values between -11.93 and -11.152 values between -10.37 and -9.591 values between -9.59 and -8.811 values between -8.81 and -8.031 values between -6.46 and -5.683 values between -4.90 and -4.122 values between -4.12 and -3.348 values between -3.34 and -2.569 values between -2.56 and -1.7823 values between -1.78 and -1.00690 values between -1.00 and -0.223046 values between -0.22 and 0.5780 values between 0.57 and 1.3515 values between 1.35 and 2.136 values between 2.13 and 2.915 values between 2.91 and 3.694 values between 3.69 and 4.472 values between 4.47 and 5.251 values between 5.25 and 6.031 values between 6.03 and 6.822 values between 6.82 and 7.601 values between 8.38 and 9.161 values between 9.94 and 10.721 values between 15.41 and 16.19offnegativepositiveoften

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. mtree
  2. boarding
  3. napshot
  4. Double
  5. idden
  6. Double
  7. .hit
  8. recht
  9. .Hit
  10. statuses

Suppresses

  1. VED
  2. boundary
  3. iota
  4. ài
  5. swell
  6. illus
  7. pear
  8. urus
  9. 舌头
  10. Head