Neuron Explorer

Qwen3-0.6B

'transistor' and electronics terms

A neuron in layer 3 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 responds to electronics words, especially the piece 'ist' inside 'transistors' and names like 'Kirchhoff'. It is strongly pushed negative by 'transistors', so that word gives the largest negative values. When turned up, the text drifted, but turning it down produced text about circuits and transistors, which fits the 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

    With a little bit of patience and dedication, you'll soon start to recognize the patterns and symbols and become an expert in no time.⏎A Brief On Kirchhoff S Laws With Circuit Diagram⏎Circuit Topology And Laws Applied Electricity⏎S

  2. #2 peak

    works but as the question asked for a simple answer I'll skip the math, google Maxwell's Equations and how they are used in the derivation of Kirchhoff's voltage law.⏎Batteries do make a good example for this simply because

  3. #3 peak

    node numbers:⏎The two large-value resistors Rbogus1 and Rbogus1 are connected across the capacitors to provide a DC path to ground in order that SPICE<|endoftext|>Writing essays The essay writing sentence

  4. #4 peak

    efficient strategies for using renewable sources of electricity to produce energy-dense liquid fuels from carbon dioxide (CO2) and biomass. Researchers will develop new families of electrocatalysts for the electrohydrogenation of CO

  5. #5 peak

    gardens near forests or villages all over Maharashtra state, he advises to those who are eager to see Mumbai butterflies to go look for them in Maharashtra Nature Park, Sanjay Gandhi National Park, Yeoor Hills in Thane or in the Vasai fort

  6. #6 peak

    to liquid fuels is a critical frontier in the transition to a renewable energy economy. One potentially game-changing technique is electrohydrogenation – the use of an electrocatalyst to drive the chemical reaction between a molecule and hydrogen.⏎The goal of

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: a pnp in the configuration. The current has a high resistance (R_H) and the gate of the transistor is connected to the ground. The gate of the transistor is connected to the ground. The gate of the transistor is connected to

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

    chip performance—measured by the number of transistors on a chip—doubles every 18 months.⏎It is no

  2. #2 peak

    Livermore National Laboratory researchers said their new tool enhances the way computer chips, LEDs and transistors are build.Commonly used piezo-electric speakers, such as those found

  3. #3 peak

    Class of products known as thermionic valves, or ‘tubes’. Valves were relegated to specialist applications once the semiconductor transistor came to dominate. Compared to transistors, valves were oversized, generated too much heat and required multiple high-voltage

  4. #4 peak

    a circuit.⏎From Tubes to Transistors⏎Basic computers that could do the same thing as the microprocessor have been

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. L0:178 suppresses
  2. L0:1224 suppresses
  3. L0:1301 suppresses
  4. L1:2801 excites
  5. L2:2328 weakly suppresses
  6. L2:75 weakly suppresses
  7. L2:41 weakly suppresses
  8. L2:111 weakly suppresses
  9. L1:394 weakly excites
  10. L2:72 weakly suppresses
  11. L2:2698 weakly excites
  12. L2:484 weakly suppresses
  13. L0:2412 weakly suppresses
  14. L2:70 weakly excites
  15. L2:1657 weakly excites
  16. L2:2620 weakly excites
  17. L2:94 weakly suppresses
  18. L2:2757 weakly suppresses
  19. L2:1494 weakly suppresses
  20. L2:1321 weakly suppresses
  21. L2:1710 weakly suppresses
  22. L2:126 weakly excites
  23. L1:1272 weakly excites
  24. L1:2683 weakly excites

Feeds later layers

  1. L7:2233 suppresses
  2. L5:1818 suppresses
  3. L4:2318 weakly suppresses
  4. L7:1439 weakly suppresses
  5. L4:2 weakly suppresses
  6. L25:1442 weakly suppresses
  7. L19:245 weakly suppresses
  8. L10:1713 weakly suppresses
  9. L8:2736 weakly excites
  10. L6:1582 weakly suppresses
  11. L15:2173 weakly suppresses
  12. L6:434 weakly suppresses
  13. L5:2042 weakly suppresses
  14. L8:2474 weakly excites
  15. L20:2655 weakly suppresses
  16. L10:1787 weakly excites
  17. L22:2645 weakly suppresses
  18. L11:1957 weakly excites
  19. L8:2129 weakly suppresses
  20. L8:1781 weakly excites
  21. L16:2137 weakly excites
  22. L23:2116 weakly excites
  23. L7:2907 weakly suppresses
  24. L8:3044 weakly 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 -2.36 and -2.292 values between -2.15 and -2.071 values between -1.64 and -1.561 values between -1.05 and -0.982 values between -0.76 and -0.691 values between -0.69 and -0.622 values between -0.62 and -0.543 values between -0.54 and -0.478 values between -0.40 and -0.3326 values between -0.33 and -0.2569 values between -0.25 and -0.18225 values between -0.18 and -0.11701 values between -0.11 and -0.031475 values between -0.03 and 0.04776 values between 0.04 and 0.11369 values between 0.11 and 0.18153 values between 0.18 and 0.2660 values between 0.26 and 0.3323 values between 0.33 and 0.409 values between 0.40 and 0.481 values between 0.48 and 0.55offnegativepositiveoften

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. ssp
  2. '})⏎⏎
  3. nak
  4. ness
  5. '$
  6. ⏎⏎
  7. specials
  8. '))⏎⏎
  9. '|
  10. 拉斯

Suppresses

  1. setActive
  2. /MIT
  3. ürger
  4. 首富
  5. TextInput
  6. enstein
  7. roi
  8. 奥巴
  9. başı
  10. 饯