Neuron Explorer

Qwen3-0.6B

'of' after 'pattern' or 'patterns'

A neuron in layer 2 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 'of' when it follows the word 'pattern' or 'patterns', as in 'a pattern of behaviour'. Turning it up did not produce a visible change in the model's replies.

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

    Washington (1994) contends that Tarquin, understood to be the poem's villain, serves to emphasize a complex pattern of meaning at work in The Rape of Lucrece. Washington maintains that Lucre

  2. #2 peak

    any element with a dark current, so to speak, will consistently report an exposure value.⏎In fact, with a long enough exposure to darkness, a disturbing pattern of exposure will appear from any CCD, because CCD elements tend to leak current. They

  3. #3 peak

    how information makes people perceive the world. Authors in many fields have noted patterns of communication that create distrust and enforce dependency by emphasizing danger from external, uncontrollable forces

  4. #4 peak

    considering both neural and molecular levels. At the neural level we can think of all of these concepts that people apply to themselves and others as patterns of firing in groups of neurons. A sufficiently complex account of neural representations can explain how

  5. #5 peak

    encian Christians. Indeed, it is perhaps more accurate to view the Mudejars as forming a distinct subsociety separated from the larger Christian society by patterns of intragroup social and cultural behavior, but intersecting with that Christian society in

  6. #6 peak

    sophisticated level. Instead of building birds' nests, we build sky scrapers. Our minds arrange events in patterns which we call designs. This memo reflects a pattern of events arranged by intent.⏎are Causes of other Patterns of Events? Patterns of events

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 problem is a classic example of a non-integer value, which is a problem that's not a real number value. However, the problem is not a real number value because the user is thinking about it, so I need to check

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

    must consider the variety of his receivers.<|endoftext|>Money is what you pay in order to get

  2. #2 peak

    .⏎Since knowledge, thinking, and rational action are properties of the individual, since the choice to exercise his rational faculty or not depends on the

  3. #3 peak

    unexpected outcomes, such as sudden drops in temperature in certain regions. This interaction exemplifies the complexity of atmospheric dynamics and its unpredictable impact on local weather conditions. And that is what we

  4. #4 peak

    1976. One hundred twenty-four years after it was first proposed, Kenneth Appel and Wolfgang Haken cleverly reduced the problem to checking the properties of 1,936 maps by computer. The result was hundreds of pages

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. L0:1079 weakly suppresses "up" after "open"
  2. L1:853 weakly excites
  3. L1:187 weakly suppresses
  4. L1:296 weakly suppresses
  5. L0:2965 weakly excites
  6. L1:712 weakly excites
  7. L0:187 weakly suppresses
  8. L1:544 weakly excites
  9. L1:2893 weakly excites
  10. L1:2380 weakly suppresses
  11. L1:1805 weakly suppresses
  12. L1:1975 weakly excites
  13. L0:1583 weakly suppresses
  14. L1:843 weakly excites
  15. L0:1373 weakly suppresses
  16. L1:2440 weakly suppresses
  17. L0:2702 weakly excites
  18. L1:128 weakly excites
  19. L1:1273 weakly excites
  20. L0:2441 weakly suppresses
  21. L1:719 weakly suppresses
  22. L1:1571 weakly excites
  23. L1:1621 weakly excites
  24. L1:2546 weakly excites

Feeds later layers

  1. L3:222 suppresses
  2. L6:607 weakly suppresses
  3. L4:266 weakly suppresses
  4. L17:2672 weakly suppresses
  5. L20:1137 weakly suppresses
  6. L9:248 weakly suppresses
  7. L8:1151 weakly suppresses
  8. L12:1280 weakly excites
  9. L8:293 weakly excites
  10. L13:989 weakly suppresses
  11. L14:1158 weakly excites 'to' starting a purpose clause
  12. L12:2926 weakly suppresses
  13. L3:659 weakly excites
  14. L5:707 weakly excites
  15. L8:2883 weakly excites
  16. L11:2774 weakly suppresses 'least' in 'at least' hedges
  17. L27:2390 weakly excites
  18. L22:1077 weakly excites
  19. L7:1757 weakly suppresses
  20. L16:2461 weakly excites
  21. L8:1954 weakly suppresses
  22. L7:478 weakly suppresses
  23. L9:2784 weakly suppresses unclear: deity names strongly negative
  24. L6:1015 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 -0.54 and -0.483 values between -0.48 and -0.4113 values between -0.41 and -0.3432 values between -0.34 and -0.2869 values between -0.28 and -0.21241 values between -0.21 and -0.14561 values between -0.14 and -0.071117 values between -0.07 and -0.011251 values between -0.01 and 0.06394 values between 0.06 and 0.13139 values between 0.13 and 0.2041 values between 0.20 and 0.2624 values between 0.26 and 0.336 values between 0.33 and 0.408 values between 0.40 and 0.473 values between 0.53 and 0.601 values between 0.60 and 0.671 values between 0.87 and 0.941 values between 0.94 and 1.001 values between 1.47 and 1.541 values between 2.08 and 2.15offnegativepositiveoften

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. okit
  2. ativos
  3. 养
  4. отов
  5. apa
  6. OVE
  7. APA
  8. onna
  9. หมาย
  10. jual

Suppresses

  1. aminer
  2. grantResults
  3. ступ
  4. �
  5. произ
  6. honored
  7. {}".
  8. غم
  9. flix
  10. false