Neuron Explorer

Qwen3-0.6B

'all' in 'at all' phrases

A neuron in layer 4 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 most to the word 'all' in phrases like 'not at all' or 'at all'. Its peaks are fairly weak compared with the strongest neurons. Turning it up or down did not clearly change the topic of the output.

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

    he disliked the Panadavas. He also knew that Krishna always favoured the Pandavas, because they were always on the side of righteousness. If at all there was a war between

  2. #2 peak

    makes you wonder why Moses included it where he did, or at all. And even more, "Why, now is this seemingly obscure story from the

  3. #3 peak

    in their discussion of creating an entrepreneurial mindset.⏎Because entrepreneurs bring together the factors of production in order to make a profit, I actually don't think entrepreneurs are at all what we need.⏎I've heard Zhao mention that students, if they want to

  4. #4 peak

    this:⏎- Sandy's wearing a lot of makeup, but it's not at all whorish. It's actually rather well done.⏎Unfortunately, while you know

  5. #5 peak

    019, The World Health Organization issued its first-ever guidance for how much screen time children under 5 should get: not very much, and none at all for those under 1.⏎The U.N. health agency said that kids under

  6. #6 peak

    the highway, Phil finds himself trapped in Punxsutawney. He wakes up the next day, only to discover that it's not the next day at all. It's Groundhog Day all over again.⏎via Wikimedia Commons⏎For some

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: to have a more concrete answer. I need to make sure that I can get it. So, the best way is to get a good deal. So, the user is clear. But I don't know. Maybe I need to have

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

    number follows it. From UPF10 to UPF50+, each number offers a certain level of protection.It’s simple: the higher the number, the more protection you get. UPF50, for example, means that only

  2. #2 peak

    help. The sooner you receive care for your anxiety symptoms, the better. Feel free to reach out to Neuro-Behavioral Clinical Research for help

  3. #3 peak

    have a heavy clay soil. Apply less water less frequently.⏎Water 10-15 gallons per week with a slow drip. The heavier the soil, the less water it may need. Watch to see how fast the water drains and avoid

  4. #4 peak

    on average. Diastolic blood pressure (the bottom number) decreased by 6 to 7 mmHg, on average. The higher the dose, the greater the drop in blood pressure tended to be. By lowering blood pressure, Mon

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. L1:2470 weakly suppresses
  2. L2:111 weakly suppresses
  3. L2:1411 weakly suppresses
  4. L2:1076 weakly excites the 'a' after 'not' or 'just'
  5. L0:450 weakly suppresses
  6. L0:633 weakly excites
  7. L3:421 weakly suppresses
  8. L1:2827 weakly suppresses
  9. L3:1937 weakly excites
  10. L2:944 weakly suppresses
  11. L3:552 weakly suppresses
  12. L0:3070 weakly excites
  13. L1:1287 weakly excites
  14. L1:1722 weakly suppresses
  15. L3:456 weakly suppresses
  16. L2:971 weakly excites
  17. L3:2322 weakly suppresses
  18. L1:1035 weakly suppresses
  19. L3:2337 weakly suppresses
  20. L0:847 weakly suppresses
  21. L3:2013 weakly excites
  22. L1:303 weakly suppresses
  23. L1:2021 weakly excites
  24. L3:1768 weakly excites

Feeds later layers

  1. L7:747 suppresses
  2. L18:1128 weakly excites
  3. L15:2896 weakly excites
  4. L5:1701 weakly suppresses
  5. L8:1960 weakly suppresses
  6. L5:2850 weakly suppresses
  7. L5:2316 weakly suppresses
  8. L7:2495 weakly excites
  9. L11:2190 weakly suppresses
  10. L7:1277 weakly suppresses
  11. L6:1681 weakly suppresses
  12. L25:545 weakly suppresses
  13. L23:2875 weakly excites
  14. L17:199 weakly suppresses
  15. L17:1205 weakly suppresses
  16. L5:2587 weakly suppresses
  17. L18:2890 weakly suppresses
  18. L20:729 weakly excites
  19. L20:1197 weakly excites 'this' pointing at the current document
  20. L6:885 weakly suppresses
  21. L11:1522 weakly excites
  22. L14:660 weakly suppresses
  23. L17:1462 weakly suppresses
  24. L19:2903 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 -2.65 and -2.571 values between -1.12 and -1.031 values between -0.79 and -0.711 values between -0.71 and -0.632 values between -0.55 and -0.472 values between -0.47 and -0.3919 values between -0.39 and -0.3140 values between -0.31 and -0.23177 values between -0.23 and -0.15500 values between -0.15 and -0.071094 values between -0.07 and 0.021064 values between 0.02 and 0.10657 values between 0.10 and 0.18235 values between 0.18 and 0.2687 values between 0.26 and 0.3415 values between 0.34 and 0.4210 values between 0.42 and 0.502 values between 0.50 and 0.58offnegativepositiveoften

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. 加工厂
  2. ception
  3. junction
  4. starter
  5. .jquery
  6. adow
  7. ctype
  8. gle
  9. ál
  10. ieves

Suppresses

  1. 如意
  2. abdomen
  3. 遏制
  4. 督办
  5. swear
  6. compact
  7. Advent
  8. 不满意
  9. Compact
  10. Endian