Neuron Explorer

Qwen3-0.6B

'see' in 'let's see' and 'see if'

A neuron in layer 17 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 fires on the word "see" in phrases like "let's see" or "see if" or "see how". It is pushed negative by words about talking, such as "talked" and "speaks". Steering did not show a clear effect on 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

    6 is conjectured to be the largest power of 2 not containing a zero. This simply stated conjecture has proven itself to be proof-resistant. Let us see why.⏎What is the probability that the nth power of two will not have any

  2. #2 peak

    - The longest beard ever was recorded at 17.4 feet long.⏎If you’re not already rocking a beard, you should try growing one out and see how you like it! However, if you don’t feel like having your own

  3. #3 peak

    . I'll see her tomorrow.⏎In #1 "going to" implies a previous arrangement or understanding

  4. #4 peak

    your kids can make a temporary lava lamp out of household items. Let’s see what you’ll need and how you can do it!⏎- Bottle (an empty

  5. #5 peak

    they recognise the parent’s voices. So reading to your baby for a few minutes gives them the comfort of your voice and listening abilities of the language.⏎Let’s see, as a parent how can we bring a change, which can make reading a

  6. #6 peak

    to as TBI, is thought to be extremely difficult to recover from. Well how about we inject some “female hormones” and see what happens?⏎In a recent article, Rita Rubin writes of an individual who suffered

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 need to not forget the need to not forget the need to not forget the need to not forget the need to not forget the need to not forget the need to not forget the need to not forget the need to not forget the need 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

    IN, MS, NC, PA, RI, TN).⏎Yet again, these numbers underscore the importance of federal and state programs for low-income people. We’ve talked about this before,

  2. #2 peak

    ECSEL classroom loves implementing is one inspired by a TEDTalk by Louie Schwartzberg, a nature time-lapse photographer. In his TEDTalk, Schwartzberg speaks about his perspective as

  3. #3 peak

    about goat feed is “can goats eat strawberries”. There is much we can talk about this topic, hence, let’s spend some time reading the article and find

  4. #4 peak

    Graham Patrick manages this with a skilful blend of necessary detail and exciting descriptions of the possibilities of current and future organic synthesis.⏎Organic chemists like to talk in their own way using a peculiar language of structures and diagrams. While this is

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. L16:1156 excites
  2. L5:1949 excites
  3. L10:1941 suppresses
  4. L10:388 weakly suppresses
  5. L16:973 weakly suppresses
  6. L4:2401 weakly suppresses
  7. L15:1688 weakly excites
  8. L16:1808 weakly suppresses
  9. L16:684 weakly excites
  10. L14:1692 weakly suppresses
  11. L10:2833 weakly excites
  12. L11:1687 weakly suppresses
  13. L13:1758 weakly suppresses
  14. L0:443 weakly suppresses
  15. L14:1024 weakly suppresses
  16. L5:448 weakly excites
  17. L15:2073 weakly excites
  18. L2:1378 weakly excites
  19. L15:2465 weakly suppresses
  20. L13:1254 weakly suppresses
  21. L14:611 weakly suppresses
  22. L4:2399 weakly suppresses
  23. L12:782 weakly excites
  24. L4:362 weakly suppresses

Feeds later layers

  1. L21:1176 weakly excites
  2. L19:1913 weakly excites
  3. L19:1177 weakly suppresses
  4. L20:766 weakly excites
  5. L23:542 weakly suppresses
  6. L27:1316 weakly excites
  7. L19:359 weakly excites
  8. L18:1489 weakly suppresses
  9. L23:33 weakly excites
  10. L19:2724 weakly excites
  11. L20:1015 weakly excites
  12. L20:2017 weakly excites
  13. L21:1016 weakly excites
  14. L20:2115 weakly excites
  15. L23:1387 weakly excites
  16. L22:1714 weakly excites
  17. L20:414 weakly suppresses
  18. L25:700 weakly suppresses
  19. L22:1085 weakly excites
  20. L22:2954 weakly excites
  21. L26:1605 weakly excites
  22. L25:2664 weakly excites
  23. L21:2117 weakly suppresses
  24. L19:945 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 -10.63 and -10.193 values between -10.19 and -9.751 values between -9.30 and -8.863 values between -7.53 and -7.091 values between -7.09 and -6.642 values between -6.64 and -6.201 values between -6.20 and -5.765 values between -5.76 and -5.313 values between -5.31 and -4.874 values between -4.87 and -4.432 values between -4.43 and -3.987 values between -3.98 and -3.545 values between -3.54 and -3.104 values between -3.10 and -2.6512 values between -2.65 and -2.2114 values between -2.21 and -1.7715 values between -1.77 and -1.3234 values between -1.32 and -0.8890 values between -0.88 and -0.441722 values between -0.44 and 0.011636 values between 0.01 and 0.45214 values between 0.45 and 0.9059 values between 0.90 and 1.3432 values between 1.34 and 1.7820 values between 1.78 and 2.233 values between 2.23 and 2.677 values between 2.67 and 3.111 values between 3.11 and 3.562 values between 3.56 and 4.002 values between 4.00 and 4.442 values between 4.44 and 4.891 values between 6.66 and 7.10offnegativepositiveoften

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. 'image
  2. 'o
  3. chner
  4. consulted
  5. ठ
  6. kn
  7. ken
  8. 的答案
  9. OPS
  10. 침

Suppresses

  1. 周
  2. anic
  3. Multiplicity
  4. 征
  5. anager
  6. width
  7. atz
  8. .prevent
  9. esi
  10. 件事