Neuron Explorer

Qwen3-0.6B

reference or starting point

A neuron in layer 20 of Qwen3-0.6B. Its strongest snippets are fairly consistent: several of them peak on the same token. It is featured, so it has steering outputs.

What this neuron does

This neuron fires on words like 'starting', 'the', or 'are' when the sentence is about a baseline, a zero point, or a place to begin, as in 'our starting point' or 'as the origin'. Pushing it up made the model talk about reference points and zero levels. The negative side is about timing words like 'shortly'.

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

    kid is, so long as we are going on-pace through the material and through the content, and reaching those educational standards and outcomes, because that’s our starting point.⏎“Our core question is, What’s the experience we want kids to have

  2. #2 peak

    by the distance from the center of the press where the forces act. Using the center of the press as the origin, distances to the left would have negative values while the distances to right would

  3. #3 peak

    glycemic index comes into the scene. Glycemic index (GI) classifies foods based on how quickly they elevate blood sugar levels, taking pure glucose as the comparison point.⏎Foods with a high GI, such as white bread and white

  4. #4 peak

    use to set goals, taking the time to plan what you want to achieve can make the difference to obtain what you want. Sort of if you know where you are going, you’ll get there. Otherwise, you may end up in a place

  5. #5 peak

    success welding galvanized tube using E70C-6 metal cored wire such as Hobart’s Galvacor; the parameters given above are a good starting point for metal cored wire. Others have found that self-shielded flux

  6. #6 peak

    of geography and the environment, who specializes in dendroclimatology.⏎“This also helps put climate change into context, as it is hard to know where you are going if you do not know where you’ve

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 a program that allows small programs (less than 1/4 of a quarter of a test case) before the program is actually running. We are running a small program that is under test with a program before the test case is

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

    candle long enough. If you light your candle and extinguish it shortly thereafter tunneling will occur. You need to burn your candle for 1 hour

  2. #2 peak

    in his body.⏎If you or anyone else has a stroke, pay special attention to the time when the symptoms started. Some treatment options are most effective when presented shortly after stroke starts.⏎Stroke signs and symptoms include:⏎Trouble talking and understanding what

  3. #3 peak

    : it will close at midnight on November 1st, shortly before the real election day on November 4. But already more than 4

  4. #4 peak

    he followed her when she returned to her home in California, and the couple were married at San Francisco in 1880. They travelled together to Europe shortly afterwards, and lived variously at London, Bournemouth, Davos and elsewhere

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. L12:2003 excites
  2. L19:2255 excites
  3. L19:2794 excites
  4. L19:2005 suppresses
  5. L15:2410 weakly suppresses
  6. L10:2896 weakly suppresses
  7. L12:2192 weakly suppresses
  8. L19:3042 weakly suppresses
  9. L8:1778 weakly suppresses
  10. L18:877 weakly suppresses
  11. L8:1888 weakly suppresses
  12. L6:1676 weakly excites
  13. L10:962 weakly suppresses
  14. L19:610 weakly excites
  15. L7:246 weakly suppresses
  16. L14:2020 weakly suppresses
  17. L7:2751 weakly excites
  18. L10:2737 weakly suppresses
  19. L4:1081 weakly excites
  20. L17:1334 weakly suppresses
  21. L9:1866 weakly excites
  22. L8:2489 weakly excites
  23. L13:2564 weakly excites
  24. L2:1440 weakly excites

Feeds later layers

  1. L22:1177 suppresses
  2. L21:2129 suppresses
  3. L27:2562 excites
  4. L24:1099 excites
  5. L23:1295 weakly excites
  6. L24:1318 weakly excites
  7. L21:3029 weakly excites
  8. L22:638 weakly excites
  9. L21:1595 weakly excites
  10. L21:2115 weakly excites
  11. L23:1703 weakly suppresses
  12. L27:2815 weakly excites
  13. L21:2726 weakly suppresses
  14. L21:2086 weakly suppresses
  15. L22:2643 weakly suppresses
  16. L23:1552 weakly suppresses
  17. L23:2872 weakly suppresses
  18. L22:2989 weakly suppresses
  19. L22:2003 weakly suppresses
  20. L22:2466 weakly excites
  21. L27:2406 weakly excites
  22. L25:1422 weakly excites
  23. L21:1350 weakly suppresses
  24. L21:161 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.

2 values between -12.39 and -11.851 values between -11.30 and -10.761 values between -8.04 and -7.501 values between -7.50 and -6.951 values between -5.32 and -4.781 values between -4.23 and -3.694 values between -3.69 and -3.145 values between -3.14 and -2.606 values between -2.60 and -2.0610 values between -2.06 and -1.5126 values between -1.51 and -0.97412 values between -0.97 and -0.422525 values between -0.42 and 0.12782 values between 0.12 and 0.6678 values between 0.66 and 1.2122 values between 1.21 and 1.758 values between 1.75 and 2.308 values between 2.30 and 2.844 values between 2.84 and 3.382 values between 3.38 and 3.934 values between 3.93 and 4.471 values between 4.47 and 5.021 values between 5.02 and 5.561 values between 6.10 and 6.651 values between 7.19 and 7.741 values between 8.82 and 9.37offnegativepositiveoften

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. zero
  2. Zero
  3. level
  4. reference
  5. 参照
  6. level
  7. 零
  8. zero
  9. 参考
  10. Level

Suppresses

  1. enting
  2. uchsia
  3. ABOVE
  4. ahan
  5. ущ
  6. 顿
  7. <Document
  8. 스트
  9. ocha
  10. getStyle