Neuron Explorer

Qwen3-0.6B

perpendicular and geometry terms

A neuron in layer 22 of Qwen3-0.6B. Its strongest snippets are fairly consistent: several of them peak on the same token. It is featured, so it has steering outputs.

What this neuron does

This neuron responds to the word 'perpendicular' and to nearby geometry vocabulary like lines, angles and slopes. When it was turned down, the model's answers drifted toward maths-style wording, though the links are loose.

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

    circle touches the larger circle and. . . .⏎X is a moveable point on the hypotenuse, and P and Q are the feet⏎of the perpendiculars from X to the sides of a right angled⏎triangle. What position of

  2. #2 peak

    's square.⏎Find the area of the shaded region created by the two overlapping⏎triangles in terms of a and b?⏎Position the lines so that they are perpendicular to each other.⏎What can you say about the equations of perpendicular lines?⏎AB

  3. #3 peak

    are parallel, perpendicular or neither?⏎Parallel lines are the lines that never intersect and perpendicular lines are the lines that intersect at 90 degrees.⏎Answer: Lines with the same slope are parallel

  4. #4 peak

    slopes are m = 2 and m = –1/2⏎If the slope of one line is the negative reciprocal of the second line, then lines are perpendicular.⏎Hence, both the lines are perpendicular.

  5. #5 peak

    two perpendicular components namely, IRMS cos φ along VRMS and IRMS sin φ perpendicular

  6. #6 peak

    relieve the symptoms of the early response, but they have little effect on late stage.⏎The non-bronchodilating or anti-inflammatory drugs such as cromolyn (Intal) and cortisone do not relieve symptoms once they occur

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 sidec c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c

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

    their pets are loving and unselfish! Share this finding with your dog-loving friends!<|endoftext|>How to Graph a Cotangent Function⏎The parent graphs of tangent and cotangent are comparable because they both have asymptotes and x-intercepts. The

  2. #2 peak

    parent graph of the cotangent function f(x) = cot x,⏎by using the same techniques you use to find the tangent parent graph

  3. #3 peak

    , or isosceles.⏎a. A triangle with no congruent sides is called __?

  4. #4 peak

    the expressions of V(t), I(t), the trigonometric formula sin(X)sin(Y)=1/2(cos(X-Y)-cos(X+Y)), and the fact that (Vmax

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. L20:1515 excites
  2. L6:946 suppresses
  3. L20:2626 excites angle and right-angle phrases
  4. L10:1571 suppresses
  5. L8:2674 weakly excites
  6. L21:2877 weakly suppresses
  7. L9:2117 weakly suppresses
  8. L17:2890 weakly excites
  9. L18:408 weakly excites
  10. L5:2452 weakly suppresses
  11. L6:779 weakly suppresses
  12. L3:187 weakly excites
  13. L11:1909 weakly excites
  14. L14:2952 weakly excites
  15. L20:1438 weakly suppresses
  16. L21:2408 weakly suppresses
  17. L9:1776 weakly suppresses
  18. L6:1814 weakly suppresses
  19. L4:2016 weakly excites
  20. L4:660 weakly excites
  21. L11:2695 weakly excites energy and resource nouns
  22. L17:488 weakly suppresses
  23. L4:540 weakly suppresses
  24. L10:2411 weakly excites

Feeds later layers

  1. L26:1055 suppresses
  2. L25:1466 suppresses
  3. L23:333 weakly suppresses
  4. L23:2554 weakly excites
  5. L27:495 weakly suppresses
  6. L24:676 weakly suppresses
  7. L24:502 weakly suppresses
  8. L23:2459 weakly suppresses
  9. L27:1833 weakly suppresses
  10. L26:1146 weakly suppresses
  11. L26:1427 weakly suppresses
  12. L23:801 weakly suppresses
  13. L23:450 weakly suppresses unclear: raccoon name fragments
  14. L24:1626 weakly suppresses
  15. L24:2393 weakly suppresses
  16. L24:714 weakly suppresses
  17. L24:1869 weakly suppresses
  18. L25:302 weakly suppresses
  19. L24:2622 weakly excites
  20. L23:941 weakly suppresses
  21. L23:741 weakly suppresses
  22. L25:3011 weakly suppresses
  23. L25:1154 weakly suppresses
  24. L27:917 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 -6.38 and -5.941 values between -4.16 and -3.712 values between -2.38 and -1.949 values between -1.49 and -1.0585 values between -1.05 and -0.60710 values between -0.60 and -0.162977 values between -0.16 and 0.2995 values between 0.29 and 0.7311 values between 0.73 and 1.176 values between 1.17 and 1.624 values between 1.62 and 2.061 values between 2.06 and 2.511 values between 2.51 and 2.951 values between 2.95 and 3.391 values between 6.06 and 6.501 values between 6.50 and 6.952 values between 10.95 and 11.39offnegativepositiveoften

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. mdi
  2. .MESSAGE
  3. ylene
  4. rale
  5. ories
  6. 瓦
  7. 凰
  8. /mark
  9. ranking
  10. 皱纹

Suppresses

  1. cott
  2. c
  3. side
  4. sides
  5. )c
  6. _side
  7. сон
  8. ไซ
  9. side
  10. SIDE