Neuron Explorer

Qwen3-0.6B

the digit 1 in COVID-19

A neuron in layer 5 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 in the word COVID-19, which is the token that comes right after 'COVID-'. It barely reacts to anything else. Turning it up or down changed the model's writing in ways that have nothing to do with COVID, so the steering results don't reveal much.

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

    facilitating a learning environment, personal growth and identifying students needing attention. Let’s discuss why healthy learning should be brought together for the betterment of students.⏎COVID-19 Has Transformed Traditional Learning Methodologies⏎Among the challenges brought by the pandemic

  2. #2 peak

    will have come up with a workable topic for their essay and learned to focus on points in their lives, rather than the entirety of it.<|endoftext|>COVID-19 Vaccine Enters Final Test Phase⏎The COVID-19 pandemic seemingly has

  3. #3 peak

    from COVID-19 is by using an alcohol-based massage or washing your hands frequently, and not touching your mouth, nose, and eyes.⏎The COVID-19 virus spreads mainly through dewdrops of saliva or discharge from the nose when the

  4. #4 peak

    Hopkins researchers recently received a $195,000 Rapid Response Research grant from the National Science Foundation to, using machine learning, identify which COVID-19 patients are at risk of adverse cardiac events such as heart failure, sustained abnormal

  5. #5 peak

    COVID-19 diagnostic based on micro/nano technology and how to tackle the above-mentioned

  6. #6 peak

    COVID-19. This is an old drug for the new virus. At least since

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 first is a simple approach of the first, which is the recursive approach. So, I think, the problem with the recursion, and the problem with the list, and the problem with the list, and the problem with the list,

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

    012, https://www.hsph.harvard.edu/nutritionsource/vitamins/.⏎- Betta Health, VIC Government, https://www.betterhealth.vic.gov.au/health/healthyliving/antioxidants⏎

  2. #2 peak

    Enchanted Learning website for:|⏎Copyright ©2002-2016

  3. #3 peak

    a highly promising new approach for the development of drugs to treat pneumococci.⏎As they report in the journal Angewandte Chemie, the scientists copied the choline architecture of the pneumococ

  4. #4 peak

    95% of all workloads will be completed in the cloud.⏎Cloud computing initiates a grand shift from the old traditional IT management systems to new and improved digitally-powered administration. Some benefits of cloud computing include:⏎- Reduction of capital costs

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. L3:270 excites
  2. L4:2141 excites
  3. L4:2286 excites
  4. L4:1362 excites
  5. L4:2046 weakly suppresses
  6. L4:2342 weakly excites
  7. L4:637 weakly suppresses
  8. L4:2240 weakly suppresses
  9. L1:1433 weakly excites
  10. L1:2789 weakly excites
  11. L1:2622 weakly excites
  12. L4:2379 weakly suppresses
  13. L4:2541 weakly excites
  14. L1:2760 weakly excites
  15. L3:1561 weakly excites
  16. L3:1860 weakly excites
  17. L1:191 weakly suppresses
  18. L1:645 weakly excites
  19. L4:2759 weakly suppresses
  20. L3:438 weakly suppresses
  21. L0:256 weakly excites
  22. L0:2608 weakly excites
  23. L2:1567 weakly excites
  24. L3:583 weakly excites

Feeds later layers

  1. L7:2512 suppresses
  2. L11:2747 weakly suppresses
  3. L8:1559 weakly excites
  4. L11:691 weakly excites
  5. L14:915 weakly suppresses
  6. L10:2413 weakly suppresses
  7. L7:573 weakly suppresses
  8. L27:2678 weakly suppresses
  9. L22:1279 weakly suppresses
  10. L21:50 weakly excites comma after 'So' or intro phrase
  11. L15:1069 weakly suppresses
  12. L12:1497 weakly excites
  13. L8:774 weakly excites
  14. L20:2598 weakly suppresses
  15. L8:1475 weakly excites
  16. L8:443 weakly excites
  17. L25:1538 weakly suppresses
  18. L9:916 weakly suppresses fractions and ratios in maths
  19. L16:2812 weakly suppresses
  20. L26:814 weakly excites
  21. L12:1802 weakly excites
  22. L21:1400 weakly suppresses
  23. L12:1319 weakly suppresses
  24. L7:2467 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 -1.13 and -0.957 values between -0.60 and -0.4292 values between -0.42 and -0.24878 values between -0.24 and -0.062362 values between -0.06 and 0.12515 values between 0.12 and 0.2930 values between 0.29 and 0.4710 values between 0.47 and 0.654 values between 0.65 and 0.832 values between 0.83 and 1.012 values between 1.18 and 1.361 values between 1.72 and 1.891 values between 1.89 and 2.071 values between 2.07 and 2.251 values between 5.27 and 5.451 values between 5.81 and 5.98offnegativepositiveoften

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. дор
  2. freezer
  3. marks
  4. rew
  5. /
  6. đạo
  7. asurable
  8. ành
  9. marks
  10. assage

Suppresses

  1. onica
  2. kaz
  3. oldValue
  4. 今天
  5. 今天我们
  6. tica
  7. сан
  8. ypes
  9. alta
  10. akens