Neuron Explorer

Qwen3-0.6B

'as' in 'just as'

A neuron in layer 1 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 reacts to the word 'as' when it comes straight after 'Just', as in 'Just as X, so Y'. The signal is fairly weak compared with the strongest neurons. Turning it up did not visibly change the model's writing in a related way.

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

    to maintain vitality throughout life – from childhood through adulthood. Just as getting older brings changes in career, family and activity goals, it also brings physiological

  2. #2 peak

    obvious way to protect your eyes from the blazing Delhi sun. Just as you use a sunscreen with adequate SPF, you must use sunglasses which offer appropriate UV

  3. #3 peak

    in the Western world, was also beset by civil war at the time. Just as Cleopatra was preparing to attack her brother with a large Arab army, the

  4. #4 peak

    there a certain kind of green zealot insists that methane may be better than coal but rather than making a prudent transition we should leap into the freezing dark. Just as so many of them are irrationally hostile to the extremely low GHG,

  5. #5 peak

    old framework which the reactionary force always tries to⏎Just as revolution is inevitable in politico-economic relations, revolution is inevitable in culture. A

  6. #6 peak

    value — the analogies are in many ways quite valid, because there are definite parallels between computational and biological viruses, and the means of battling against them. Just as we humans can minimize the risks of ill health through prevention and curing, the same

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 been thinking about the impact of the world. I have been thinking about the impact of the world. I have been thinking about the impact of the world. I have been thinking about the impact of the world. I have been thinking

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

    overdoses, they tried to use newer forms of opioids to treat the addiction to its

  2. #2 peak

    ens but this study shows that it may be more complicated than that - asthma may involve colonization of the airways by multiple bacteria," said study co-author Homer Boushey, MD, a UCSF professor of medicine in the division of Pulmonary

  3. #3 peak

    why is it that when confronted with compelling evidence that college pays a big financial dividend, so many young people still don’t get a college degree?⏎Heather Boushey and I argue that the short answer is that for a surprising share of college

  4. #4 peak

    the policy of coexistence has been implemented in the past ten years.<|endoftext|>BBB - A+ Rating - the best there is⏎An excerpt from www.HouseOfNames.com archives copyright © 2000 - 201

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. L0:1301 strongly suppresses
  2. L0:178 strongly suppresses
  3. L0:1224 suppresses
  4. L0:498 suppresses
  5. L0:397 weakly suppresses
  6. L0:41 weakly excites
  7. L0:760 weakly suppresses
  8. L0:1 weakly suppresses
  9. L0:2252 weakly suppresses
  10. L0:1036 weakly excites
  11. L0:1010 weakly suppresses
  12. L0:2518 weakly suppresses
  13. L0:962 weakly suppresses
  14. L0:1388 weakly suppresses
  15. L0:497 weakly suppresses
  16. L0:2759 weakly suppresses
  17. L0:1947 weakly suppresses
  18. L0:93 weakly suppresses
  19. L0:953 weakly suppresses
  20. L0:799 weakly suppresses
  21. L0:2944 weakly suppresses
  22. L0:2497 weakly suppresses
  23. L0:162 weakly suppresses
  24. L0:1063 weakly suppresses

Feeds later layers

  1. L12:129 weakly excites unclear: mixed, mostly everyday health and advice prose
  2. L2:400 weakly excites
  3. L11:622 weakly excites
  4. L18:2593 weakly suppresses
  5. L8:803 weakly excites
  6. L11:1670 weakly excites
  7. L12:1825 weakly excites
  8. L2:2831 weakly suppresses
  9. L11:2812 weakly excites
  10. L13:1847 weakly excites
  11. L20:376 weakly excites
  12. L14:1945 weakly suppresses
  13. L6:2215 weakly suppresses
  14. L11:1786 weakly excites
  15. L15:494 weakly excites
  16. L21:2623 weakly suppresses
  17. L12:567 weakly suppresses
  18. L6:2820 weakly suppresses
  19. L24:2112 weakly suppresses
  20. L26:370 weakly excites
  21. L9:462 weakly excites
  22. L2:2617 weakly excites
  23. L6:2700 weakly suppresses
  24. L3:118 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 -0.53 and -0.501 values between -0.50 and -0.481 values between -0.48 and -0.452 values between -0.45 and -0.424 values between -0.42 and -0.401 values between -0.40 and -0.374 values between -0.37 and -0.357 values between -0.35 and -0.3210 values between -0.32 and -0.2915 values between -0.29 and -0.2717 values between -0.27 and -0.2434 values between -0.24 and -0.2150 values between -0.21 and -0.1965 values between -0.19 and -0.1696 values between -0.16 and -0.14133 values between -0.14 and -0.11213 values between -0.11 and -0.08308 values between -0.08 and -0.06465 values between -0.06 and -0.03581 values between -0.03 and -0.01513 values between -0.01 and 0.02358 values between 0.02 and 0.05246 values between 0.05 and 0.07189 values between 0.07 and 0.10141 values between 0.10 and 0.13121 values between 0.13 and 0.1592 values between 0.15 and 0.1867 values between 0.18 and 0.2055 values between 0.20 and 0.2337 values between 0.23 and 0.2625 values between 0.26 and 0.2815 values between 0.28 and 0.318 values between 0.31 and 0.339 values between 0.33 and 0.367 values between 0.36 and 0.395 values between 0.39 and 0.412 values between 0.41 and 0.443 values between 0.44 and 0.464 values between 0.46 and 0.493 values between 0.49 and 0.52offnegativepositiveoften

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. amental
  2. =val
  3. antes
  4. naments
  5. etting
  6. mel
  7. ackers
  8. synthesize
  9. cis
  10. imitives

Suppresses

  1. eki
  2. under
  3. seeds
  4. hti
  5. breadcrumbs
  6. litter
  7. TL
  8. EPA
  9. 阱
  10. Majesty