Neuron Explorer

Qwen3-0.6B

'to' after 'too big' or 'right angles'

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 to the word 'to' when it follows an adjective or a phrase like 'too big', 'too light', or 'at right angles', as in 'too big to be human'. It is most strongly pushed down by 'to' in phrases about getting somewhere, like 'all the way to the top'. Steering did not give a clear theme.

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

    defined in terms of its⏎(i) Isometric System :⏎Three axes of crystal are⏎of equal length and lie at right angle to each⏎other, e.g. Halite mineral.⏎(ii) Tetragonal

  2. #2 peak

    were satisfied with these explanations. Some radar operators and pilots insisted that the movements of the objects were too coordinated to be random atmospheric anomalies. Furthermore, skeptics pointed out that temperature inversions were

  3. #3 peak

    pure cinnabar sources. When sold in Europe, it was frequently adulterated with red lead, an inexpensive bright lead oxide pigment that was too reactive to be trustworthy enough for use in art. Vermilion was favored by old paint masters

  4. #4 peak

    see this for yourself, stick two fridge magnets together, back to back. They will slide against each other in one direction. If you slide them at right angles to that, they will alternately attract and repel one another.<|endoftext|>E.

  5. #5 peak

    .⏎A couple of days later, police and investigation officers showed up on their door, as per a local news report. They determined that the bones were too big to be human, and told the couple that it was of

  6. #6 peak

    axes is perpendicular⏎to other axes and is of different length, e.g.⏎(iv) Orthorhombic System :⏎Three axes intersect⏎at right angles to each other, but they⏎are of unequal lengths, e.g. Stau

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 can see a lot of the point of the goal, and I am able to reach it. I can feel it. I have reached the goal. I can feel it. I can feel it. I have reached the goal. I

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

    a meteor passes through the Earth's atmosphere at high speeds, it heats up, glows brightly and finally vaporizes. If a meteor makes it all the way to the ground, it is called a meteorite.⏎Meteors occur randomly during

  2. #2 peak

    . Even if you don't make it all the way to the top of the mountain or hike all 40 miles of the trail,

  3. #3 peak

    is up to, and she slams the window sash on his beard. Trapped, he must surrender the ladder.⏎After climbing to the top of the mountain, she releases him by pulling on a long rope.

  4. #4 peak

    during the germination period (usually 1 - 2⏎weeks) by applying frequent, but light, irrigations.⏎Large lawns or hard-to-reach areas may be better hydroseeded.⏎Pattern for sowing seed<|endoftext|>MONT

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. L3:1859 suppresses
  2. L2:872 excites
  3. L2:750 weakly excites
  4. L3:592 weakly suppresses
  5. L3:59 weakly excites
  6. L3:50 weakly excites
  7. L0:1456 weakly suppresses
  8. L3:2267 weakly suppresses
  9. L3:2179 weakly suppresses
  10. L2:2875 weakly excites 'up' after show, turn and showed
  11. L3:1181 weakly suppresses
  12. L3:1383 weakly suppresses
  13. L2:1919 weakly excites
  14. L2:1321 weakly suppresses
  15. L2:1174 weakly suppresses
  16. L3:272 weakly excites
  17. L3:122 weakly excites
  18. L2:2956 weakly suppresses
  19. L3:115 weakly suppresses
  20. L2:2304 weakly suppresses
  21. L1:906 weakly suppresses
  22. L3:127 weakly excites
  23. L2:1148 weakly suppresses
  24. L2:1036 weakly excites

Feeds later layers

  1. L26:1264 suppresses
  2. L8:2289 suppresses
  3. L8:1627 weakly suppresses
  4. L16:2690 weakly suppresses
  5. L7:1883 weakly suppresses
  6. L23:1277 weakly suppresses
  7. L25:1528 weakly suppresses
  8. L8:1716 weakly suppresses
  9. L17:2282 weakly suppresses
  10. L9:723 weakly suppresses
  11. L16:2744 weakly suppresses
  12. L16:2027 weakly suppresses
  13. L26:676 weakly excites
  14. L26:3055 weakly suppresses
  15. L5:1903 weakly excites
  16. L16:1168 weakly suppresses 'to' after verbs of change
  17. L13:2597 weakly suppresses
  18. L23:2455 weakly suppresses
  19. L11:2700 weakly excites
  20. L17:1369 weakly suppresses "of" after is/are/be (of + quality)
  21. L6:1149 weakly suppresses
  22. L10:2980 weakly excites
  23. L7:2555 weakly excites
  24. L14:1599 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 -5.85 and -5.661 values between -4.73 and -4.541 values between -4.17 and -3.981 values between -3.42 and -3.231 values between -2.49 and -2.301 values between -2.11 and -1.932 values between -1.93 and -1.742 values between -1.74 and -1.551 values between -1.55 and -1.361 values between -1.36 and -1.183 values between -0.99 and -0.805 values between -0.80 and -0.625 values between -0.62 and -0.4361 values between -0.43 and -0.24755 values between -0.24 and -0.062471 values between -0.06 and 0.13539 values between 0.13 and 0.3235 values between 0.32 and 0.509 values between 0.50 and 0.694 values between 0.69 and 0.881 values between 0.88 and 1.075 values between 1.07 and 1.251 values between 1.25 and 1.442 values between 1.44 and 1.63offnegativepositiveoften

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. visto
  2. izes
  3. dụ
  4. Knife
  5. Wrapper
  6. Int
  7. fires
  8. Wrapper
  9. гон
  10. قر

Suppresses

  1. 设计理念
  2. reached
  3. reach
  4. 抵达
  5. 到达
  6. reachable
  7. reach
  8. reaching
  9. .sf
  10. 终点