Neuron Explorer

Qwen3-0.6B

the word tibia (shin bone)

A neuron in layer 5 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 on the word tibia, the shin bone, and on a few other bone names. When turned up, the model's writing wandered without any clear link to bones, so only the text examples support this.

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

    any known injury at that time. Examination reveals full motion of the right hip, knee, and ankle. He has tenderness to palpation over the anterior tibia with minimal swelling. No erythema

  2. #2 peak

    works was the Libellus de abbreviaturis Hebraeorum, tam Talmudicorum quam Masoritarum et aliorum rabbinorum (Paris, 1561), later exploited by Guy *Le

  3. #3 peak

    one of the major ligaments of the knee that is located in the middle of the knee and runs from the femur (thigh bone) to the tibia (shin bone). It prevents the tibia from sliding out in front of

  4. #4 peak

    20 years of age, and stabilizes when the bones are no longer growing.⏎Almost all OFD tumors occur in the long middle part of the tibia, with a few, rare

  5. #5 peak

    1% of all cancerous tumors that begin in the bone are adamantinomas.⏎These slow growing tumors most often occur in the middle part of the tibia, and in many cases, the fibula bone is also affected. Although uncommon

  6. #6 peak

    distal tibial shaft with no significant swelling. Radiographs are shown in Figure A and B. What is the most appropriate treatment?⏎MRI of the tibia⏎Aspiration of the tibia⏎Referral to child services⏎Long leg

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 1000.000.000.000.000.000.000.000.000.000.000.0

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

    to the Hittites around 1150 BC. They had healthy relations with the Kassites and Hurrians. ⏎After the Battle of Megiddo (1121 BC) Egypt pushed Mit

  2. #2 peak

    saying - Write a review⏎Respiratory system 52 Thoracotomy⏎Stem cell transplantation⏎Cardiovascular system 56 Coronary artery bypass graft⏎Cardiac valve surgery⏎Gastrointestinal system 29 Gastro

  3. #3 peak

    in agents to conduct the land sale. In the midst of rich farmland, Whitewright soon drew settlers from the small, nearby communities that the railroad had bypassed

  4. #4 peak

    by the team. Moreover, QA is about engineering processes that assure quality is achieved in an effective and efficient way.⏎What should be on a QA tester’s résumé?⏎For instance, if a defect is found and fixed, there is no

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. L4:591 strongly excites
  2. L4:524 strongly suppresses
  3. L2:598 strongly suppresses
  4. L4:643 excites mid-word pieces of technical words
  5. L2:1853 excites
  6. L3:2273 excites
  7. L4:814 suppresses
  8. L4:345 suppresses
  9. L4:1345 suppresses
  10. L2:1091 excites
  11. L4:609 suppresses
  12. L4:1214 suppresses
  13. L0:1395 suppresses
  14. L4:1581 excites
  15. L2:761 excites
  16. L1:259 excites
  17. L4:1045 excites
  18. L2:1202 excites
  19. L2:801 excites
  20. L4:490 excites decade digit in years like 1930s
  21. L2:1969 excites
  22. L2:1253 suppresses
  23. L4:2239 suppresses
  24. L3:2921 excites

Feeds later layers

  1. L6:177 strongly suppresses
  2. L10:94 strongly excites
  3. L16:1231 strongly excites
  4. L6:853 excites
  5. L19:395 excites
  6. L22:1055 excites
  7. L21:625 excites
  8. L24:118 excites
  9. L21:327 excites
  10. L17:446 excites
  11. L25:1493 excites
  12. L15:2378 excites
  13. L18:1399 excites
  14. L16:286 excites
  15. L6:2548 excites
  16. L8:1546 excites
  17. L18:931 excites
  18. L8:423 excites
  19. L17:647 excites
  20. L22:271 suppresses
  21. L8:71 excites
  22. L15:2289 excites
  23. L6:1100 excites
  24. L19:1059 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.

4 values between -0.63 and -0.581 values between -0.58 and -0.544 values between -0.54 and -0.498 values between -0.49 and -0.4523 values between -0.45 and -0.4040 values between -0.40 and -0.36108 values between -0.36 and -0.31177 values between -0.31 and -0.27254 values between -0.27 and -0.22367 values between -0.22 and -0.18422 values between -0.18 and -0.13450 values between -0.13 and -0.09538 values between -0.09 and -0.04674 values between -0.04 and 0.00405 values between 0.00 and 0.05189 values between 0.05 and 0.0973 values between 0.09 and 0.1450 values between 0.14 and 0.1842 values between 0.18 and 0.2314 values between 0.23 and 0.2717 values between 0.27 and 0.328 values between 0.32 and 0.367 values between 0.36 and 0.419 values between 0.41 and 0.455 values between 0.45 and 0.502 values between 0.50 and 0.545 values between 0.54 and 0.591 values between 0.59 and 0.631 values between 0.63 and 0.682 values between 0.68 and 0.722 values between 0.72 and 0.771 values between 0.77 and 0.811 values between 0.81 and 0.861 values between 0.86 and 0.901 values between 0.95 and 0.991 values between 0.99 and 1.041 values between 1.13 and 1.18offnegativepositiveoften

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. theless
  2. ly
  3. ️
  4. ic
  5. gether
  6. as
  7. ов
  8. ful
  9. ся
  10. empre

Suppresses

  1. Vish
  2. plotlib
  3. //-
  4. plies
  5. 砧
  6. )[-
  7. Prosper
  8. wegian
  9. tôn
  10. .mj