Neuron Explorer

Qwen3-0.6B

'corners' in Four Corners

A neuron in layer 2 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 fires strongly on the word 'corners', especially in the phrase 'Four Corners' (the US region where four states meet). Turning it up did not make the writing mention corners, so the effect on output is unclear.

Written by Claude from this neuron's activations and steering outputs. Confidence: high. 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

    they have, be a good example to them.⏎#9. Keep at It.⏎Talking about mone<|endoftext|>Next: My Tiles are Cracking in the Corners⏎My tiles are wearing out⏎Tiles are well known for their durability and wear

  2. #2 peak

    Utes have natural gas, the Navajo Nation has coal, which has for decades been extracted to feed various large-scale power plants, including in the Four Corners. But that revenue has come under threat by tightening air-quality regulations. At the

  3. #3 peak

    Service. "The weed population appears to be increasing exponentially in certain areas of our state."⏎Invasive weeds that choke out native vegetation are creeping into the Four Corners area from the north, south, east and west. This "slow-motion explosion

  4. #4 peak

    Diffuse knapweed can spread when dried seed heads attach to vehicles, people or animals.⏎Other knapweed varieties also are major problems in the Four Corners area. Russian knapweed is the most common, overtaking fields and empty

  5. #5 peak

    monsoon rains to return to the Four Corners and Western Slope in July. But there’s still a danger in the mountains

  6. #6 peak

    ainer: What can the oil companies do to stave off internal pipe corrosion? Very little. It would be far<|endoftext|>DLTK's Crafts for Kids⏎Cone Santa Paper Craft⏎Set Santa out on a coffee table or the mantle as a

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 have been thinking about... I have been thinking about... I have been thinking about... I have been thinking about... I have been thinking about... I have been thinking about... I have been thinking about... I have been thinking about...

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

    2. Subclinical diabetes⏎Subclinical diabetes is understood when the patient is completely asymptomatic and fasting blood glucose level is usually normal but the post prandial level (glucose levels after two hours of meal) is frequently elevated.⏎

  2. #2 peak

    14). To elucidate the drought responses of these two tree species to a long-term and repeated drought in monoculture and mixed plantings, the “Kranzberg Roof Project” (KROOF, kroof.wzw.tum

  3. #3 peak

    maintaining shrubs: heading back, renewal, rejuvenation, and shearing. For information on pruning specific shrubs, download the following list.⏎TIME TO PRUNE⏎A general rule-of-thumb is that shrubs that bear flower buds on the

  4. #4 peak

    mama putro mama bhrAtA mama vatso.atha gaurmama |⏎vR^ikairvyApAditA hyevaM krandanti sma gR^ihe gR^ihe ||2-

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. L1:1110 weakly excites
  2. L1:5 weakly excites
  3. L0:178 weakly suppresses
  4. L1:2240 weakly excites
  5. L0:1224 weakly suppresses
  6. L1:1 weakly suppresses
  7. L1:1099 weakly excites
  8. L0:1301 weakly suppresses
  9. L1:23 weakly suppresses
  10. L1:99 weakly excites
  11. L1:18 weakly excites
  12. L1:1960 weakly excites
  13. L1:2118 weakly suppresses
  14. L1:331 weakly suppresses
  15. L1:4 weakly suppresses
  16. L1:538 weakly suppresses
  17. L1:1454 weakly suppresses
  18. L1:2638 weakly suppresses
  19. L0:497 weakly suppresses
  20. L1:1510 weakly suppresses
  21. L0:114 weakly suppresses
  22. L0:1632 weakly suppresses
  23. L0:1403 weakly excites
  24. L1:134 weakly excites

Feeds later layers

  1. L19:1605 weakly excites
  2. L12:1347 weakly suppresses
  3. L7:160 weakly suppresses
  4. L17:1892 weakly excites
  5. L24:1352 weakly excites
  6. L24:300 weakly suppresses
  7. L4:2944 weakly suppresses
  8. L3:2601 weakly excites
  9. L7:1081 weakly excites
  10. L13:1534 weakly excites
  11. L15:1984 weakly suppresses
  12. L9:2668 weakly excites
  13. L17:1670 weakly excites
  14. L16:2250 weakly suppresses
  15. L11:991 weakly excites
  16. L13:602 weakly suppresses
  17. L10:2060 weakly suppresses
  18. L26:1193 weakly excites
  19. L15:1792 weakly excites
  20. L4:2016 weakly excites
  21. L3:2112 weakly excites
  22. L10:2918 weakly excites
  23. L17:1605 weakly suppresses
  24. L14:2934 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 -1.04 and -0.971 values between -0.91 and -0.841 values between -0.84 and -0.771 values between -0.70 and -0.631 values between -0.56 and -0.491 values between -0.49 and -0.422 values between -0.42 and -0.3613 values between -0.36 and -0.2939 values between -0.29 and -0.22157 values between -0.22 and -0.15551 values between -0.15 and -0.081144 values between -0.08 and -0.011164 values between -0.01 and 0.06486 values between 0.06 and 0.13228 values between 0.13 and 0.1968 values between 0.19 and 0.2627 values between 0.26 and 0.3317 values between 0.33 and 0.404 values between 0.40 and 0.471 values between 0.47 and 0.541 values between 1.64 and 1.71offnegativepositiveoften

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. �
  2. face
  3. o
  4. treffen
  5. atsby
  6. 理论
  7. 面
  8. styl
  9. 扪
  10. face

Suppresses

  1. |RF
  2. lined
  3. lining
  4. Bail
  5. sendo
  6. |r
  7. \|
  8. каж
  9. |R
  10. oogle