Neuron Explorer

Qwen3-0.6B

lowercase m opening a word fragment

A neuron in layer 27 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 fires on a lowercase 'm' that begins the middle or end of a word, as in 'mussels', 'mucus' or 'molar'. Pushing it up or down changed the model's output into rambling or repeated text, which does not reveal much about the letter.

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

    population of this fascinating creature and they are currently balancing on a knife-edge.⏎Populations are found in some of the rivers flowing into Loch Ness.⏎Freshwater pearl mussels are very slow-growing and live at the bottom of clean, generally fast-flow

  2. #2 peak

    Sanskrit.⏎|karsharpan / karsha (coin)||cash|⏎|mukh (face)||Mouth (old english muth)|⏎|kapal (head)||as in per “capita”,

  3. #3 peak

    . This process is more formally known as tearing apart. Place your book on a flat surface, and with a sharp knife, cut right through the endsheet, mull and the covering material as depicted in figure 180. While doing

  4. #4 peak

    .95 million years ago rekindle debate over whether our own genus, Homo, split into multiple lineages early on.⏎- Analysis of tartar, molar wear and tooth chemistry in the nearly two-million-year-old hominin known as

  5. #5 peak

    ia and finally home to Cartage.⏎The flies we buzzing merrily around his bed. As he tiredly met the dawn, he thought

  6. #6 peak

    consequences. Egan provides a chilling account of how sea lamprey, zebra and quagga mussels and other invaders have made their way into the lakes, decimating native species

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: Whenineededineedineedineedineedineedineedineedineedineedineedineedineedineedineedineedineedineedineedineedineedineedineed

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

    : the German-Fascists invaded the territory of the country, a war began. Young boys and girls-the same age as the current students of KSMU-went to the front as soldiers, nurses, doctors, sacrificing themselves and defending their

  2. #2 peak

    The COVID-19 pandemic has worsened problems that were already bedeviling Nigeria’s health system, including run-down or inadequate medical equipment, underpaid and unmotivated medical staff, and unequal access to health care (Andolu Agency,

  3. #3 peak

    1986 because of the Chernobyl nuclear disaster; Centralia in Pennsylvania, cleared out over 20+ years due to an underground coal-mine fire that’s been burning since 1962, which has made

  4. #4 peak

    OH and South Bend, IN established all-time records for those locations. Until this date, Erie, PA, was one of the few non-mountainous spots in<|endoftext|>The New Student's Reference Work/Alcinous⏎

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. L22:1138 suppresses
  2. L24:1190 excites
  3. L21:1679 excites
  4. L17:230 excites
  5. L21:505 suppresses
  6. L22:2646 excites
  7. L20:234 suppresses
  8. L20:417 excites
  9. L20:1120 excites
  10. L20:87 excites
  11. L18:133 excites
  12. L10:311 weakly suppresses
  13. L7:922 weakly suppresses
  14. L7:291 weakly excites
  15. L19:251 weakly excites
  16. L1:333 weakly excites
  17. L16:34 weakly suppresses
  18. L19:381 weakly excites word-start 'acc' fragments
  19. L5:1642 weakly suppresses
  20. L13:2993 weakly excites
  21. L21:463 weakly suppresses
  22. L9:1834 weakly suppresses
  23. L16:255 weakly excites
  24. L0:73 weakly excites

Feeds later layers

None: this is the last layer.

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.

2 values between -19.30 and -13.371 values between -13.37 and -7.4515 values between -7.45 and -1.523878 values between -1.52 and 4.416 values between 4.41 and 10.332 values between 10.33 and 16.261 values between 22.19 and 28.112 values between 51.82 and 57.741 values between 211.82 and 217.75offnegativepositiveoften

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. UsersController
  2. Partition
  3. роб
  4. Partition
  5. behalf
  6. writeln
  7. icularly
  8. pours
  9. PLIED
  10. pping

Suppresses

  1. anch
  2. itia
  3. oras
  4. ine
  5. 带
  6. ounter
  7. Lİ
  8. ishly
  9. ix
  10. ito