Neuron Explorer

Qwen3-0.6B

single letter starting a split word

A neuron in layer 26 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

It fires on a lone letter, such as 'o' or 'f', that begins a word which is split into pieces, like 'o' in 'opossum' or 'f' in 'fidgeting'. Steering did not produce 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

    for your yard, but it’s diet of fruits and vegetables can also make it a pest that destroys your plants and any vegetable garden. The opossum keeps to itself, traveling alone, and is not a pack animal.

  2. #2 peak

    , terminal and lateral; bracts and bracteoles absent; rays numerous, upwards-spreading, unequal. Calyx teeth obsolete. Petals yellow, obovate, mid rib conspicuous, apex with narrowly inflexed lobule

  3. #3 peak

    rets with olive oil, garlic, and parmesan cheese and roast in the oven for a crispy and flavorful side dish.⏎- Broccoli and mushroom omelette: Saute broccoli florets and sliced mushrooms and fold them into a

  4. #4 peak

    from a medical condition that needs treatment.⏎Common symptoms of anxiety would be increased heart rate, hyperventilation, sweating, nervousness, weakness, fatigue, fidgeting, inability to concentrate, nausea or other stomach related issues, and difficulty

  5. #5 peak

    , parabolas, ellipses, hyperbolas, cubic curves like elliptic curves and quartic curves like lemniscates, and Cassini ovals. A point of the

  6. #6 peak

    a dog’s leg and break the ankle of a human, and the poison causes gangrene if untreated. They live along the banks of rivers and oasises.⏎They ambush creatures, hiding in the sand and snapping when a foot

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: people shouldn't be allowed forthtainingtainingtainingtainingtainingtainingtainingtainingtainingtainingtainingtainingtainingtainingtainingtainingtainingtainingtainingtainingt

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

    exploitation and human trafficking in Thailand.⏎Human trafficking has become a well-known topic and a well-loved cause during the past several years, but it still remains one of the fastest

  2. #2 peak

    tics. The effect may only last for a few months, and repeated treatments might result in tolerance, rendering the drug ineffective after several applications.⏎Treatment of Attention Deficit Hyperactive Disorder (ADHD)⏎ADHD is not uncommon in children with

  3. #3 peak

    “Interdisciplinary System for Interactive Scientific and Scientific-Technical Information”.<|endoftext|>An team of researchers heading to the Antarctic later this year may go in search of a long-lost piece of exploration history. A group of international climate scientists will travel into the

  4. #4 peak

    each chakra represents? Here is a simple diagram and list for you to follow. | WIKIPEDIA⏎THE CHAKRAS ARE AS FOLLOWS⏎- ROOT/BASE (MULADHARA) CHAKRA

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. L23:2981 suppresses
  2. L22:369 excites
  3. L16:177 excites zero digit inside dates and codes
  4. L18:437 excites
  5. L16:67 excites
  6. L17:213 suppresses
  7. L10:90 suppresses
  8. L5:1546 suppresses
  9. L2:1622 excites
  10. L18:112 suppresses 'over' meaning time or duration
  11. L0:44 suppresses
  12. L1:15 suppresses
  13. L21:381 suppresses zeros inside long number codes
  14. L2:1795 weakly suppresses
  15. L18:328 weakly suppresses
  16. L17:378 weakly excites
  17. L2:1787 weakly suppresses
  18. L5:926 weakly excites
  19. L7:2095 weakly suppresses
  20. L17:1046 weakly suppresses
  21. L20:211 weakly excites
  22. L10:1052 weakly suppresses
  23. L7:287 weakly suppresses
  24. L19:702 weakly excites

Feeds later layers

  1. L27:686 weakly suppresses
  2. L27:903 weakly suppresses
  3. L27:346 weakly suppresses
  4. L27:2684 weakly suppresses
  5. L27:210 weakly suppresses
  6. L27:455 weakly excites
  7. L27:2875 weakly excites
  8. L27:420 weakly suppresses
  9. L27:1435 weakly suppresses
  10. L27:2857 weakly suppresses
  11. L27:918 weakly suppresses
  12. L27:1427 weakly suppresses
  13. L27:2633 weakly suppresses
  14. L27:1386 weakly suppresses
  15. L27:654 weakly suppresses
  16. L27:2118 weakly excites
  17. L27:1068 weakly suppresses
  18. L27:488 weakly excites
  19. L27:493 weakly suppresses
  20. L27:652 weakly suppresses
  21. L27:2374 weakly suppresses
  22. L27:929 weakly suppresses
  23. L27:1940 weakly excites
  24. L27:1825 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 -36.06 and -34.701 values between -23.84 and -22.481 values between -21.12 and -19.761 values between -12.97 and -11.621 values between -11.62 and -10.261 values between -10.26 and -8.902 values between -8.90 and -7.542 values between -4.82 and -3.475 values between -3.47 and -2.1160 values between -2.11 and -0.753771 values between -0.75 and 0.6137 values between 0.61 and 1.978 values between 1.97 and 3.332 values between 3.33 and 4.681 values between 4.68 and 6.048 values between 6.04 and 7.401 values between 8.76 and 10.122 values between 10.12 and 11.481 values between 11.48 and 12.831 values between 14.19 and 15.551 values between 16.91 and 18.27offnegativepositiveoften

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. 郦
  2. opol
  3. oug
  4. nom
  5. lugar
  6. EFR
  7. books
  8. ynchronized
  9. ENTE
  10. ogl

Suppresses

  1. ichi
  2. aining
  3. nds
  4. t
  5. ungen
  6. te
  7. icho
  8. tool
  9. 局面
  10. aight