Neuron Explorer

Qwen3-0.6B

unclear: mid-word and descriptive phrases

A neuron in layer 14 of Qwen3-0.6B. Its strongest snippets are mixed: they peak on different tokens. It is featured, so it has steering outputs.

What this neuron does

This neuron fires on scattered spots in everyday descriptive text, such as the start of a word split across tokens or a phrase like 'a wide variety [of]'. The pattern is hard to pin down. Turning it up produced off-topic and repetitive text, so it does not reveal a clear meaning.

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

    propylene foam or formed thermoplastic.⏎What is a bumper absorber?⏎The bumper absorber is located behind the decorative plastic bumper cover. As its name implies, the bumper absorber takes up the impact energy in a collision, to protect

  2. #2 peak

    .⏎Shovels, Spades, and Trowels⏎Shovels are among the most basic and essential gardening tools. Shovels have many different-shaped blades, depending on whether they are designed for digging holes or for moving around piles

  3. #3 peak

    ; to break up hard, clumpy soil; and to remove weeds. Hoes come in many different sizes and weights; they also come in a wide variety of blade shapes, including flat, angled, stirrup and looped-style blades.

  4. #4 peak

    Sweden about every fourth person aged 45 years or older has doctor-diagnosed osteoarthritis in one or multiple joints, the knee being the most frequent location. As we have a steadily ageing and increasingly obese population in Sweden the prevalence

  5. #5 peak

    sword and slay dragons.⏎St George’s Requests⏎However, we can remember that after saving the king’s daughter, George made requests, one of which was never to forget the poor, which is Christian thinking but also transcultural

  6. #6 peak

    utras of Patañjali (200 B. C.). The word ‘ekabhavika’ has been introduced and discussed here. Literally, it means ‘producing one birth’.⏎In the residual karma

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 something I has been was, it was, was, was, was, was, was, was, was, was, was, was, was, was, was, was, was, was, was, was, was

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

    than lifting water off the Earth” such that spaceships won’t have the tremendous expense of lifting them from Earth in order to bring them along. They could theoretically travel from Earth to the lunar waypoint and find resources waiting for them before undertaking a journey

  2. #2 peak

    most are bitmapped formats -- especially hard to collaborate on -- and, worse, many incorporate data compression. Most of the rest require proprietary tools or operating systems to function, which makes collaboration across organizational boundaries difficult. Of particular importance: most vector formats

  3. #3 peak

    in the 14th century, Flemish artisans rose up in a series of bloody revolts against the aristocratic cloth dealers who had long monopolized power. The Count of Flanders and the

  4. #4 peak

    , bushy clumps. Axes extend through entire length of plant, but main⏎branches absent; dichotomous branching throughout. Axils rounded, with sides⏎symmetrically developed, with side branches at times growing almost⏎horizontally

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. L11:1576 strongly excites
  2. L12:1810 suppresses
  3. L13:1198 excites
  4. L13:2112 excites
  5. L13:1330 suppresses
  6. L13:1807 excites
  7. L12:817 excites
  8. L13:1504 excites
  9. L12:1070 excites
  10. L12:1436 suppresses
  11. L13:2422 excites
  12. L13:998 excites
  13. L12:1425 suppresses
  14. L13:602 excites
  15. L13:2768 suppresses
  16. L13:1402 excites
  17. L12:2103 suppresses
  18. L12:1445 suppresses
  19. L13:1720 suppresses
  20. L8:1609 excites
  21. L11:1442 excites
  22. L13:174 excites
  23. L12:2194 excites
  24. L13:110 excites

Feeds later layers

  1. L15:555 suppresses
  2. L15:1181 excites
  3. L15:1637 excites
  4. L15:1822 excites
  5. L15:1941 suppresses
  6. L15:1689 suppresses
  7. L16:1745 excites
  8. L15:2404 suppresses
  9. L18:451 suppresses
  10. L15:1602 suppresses
  11. L15:2136 excites
  12. L16:1108 suppresses
  13. L16:2638 excites
  14. L16:787 excites
  15. L15:149 weakly suppresses
  16. L15:534 weakly excites
  17. L17:1198 weakly suppresses
  18. L17:64 weakly excites
  19. L15:192 weakly excites
  20. L19:261 weakly suppresses
  21. L21:1540 weakly excites
  22. L15:1073 weakly suppresses
  23. L17:767 weakly suppresses
  24. L17:1402 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.

4 values between -1.58 and -1.481 values between -1.48 and -1.383 values between -1.38 and -1.282 values between -1.28 and -1.172 values between -1.17 and -1.072 values between -1.07 and -0.978 values between -0.97 and -0.875 values between -0.87 and -0.7710 values between -0.77 and -0.6618 values between -0.66 and -0.5637 values between -0.56 and -0.4659 values between -0.46 and -0.3688 values between -0.36 and -0.26149 values between -0.26 and -0.16454 values between -0.16 and -0.051631 values between -0.05 and 0.05722 values between 0.05 and 0.15346 values between 0.15 and 0.25160 values between 0.25 and 0.3582 values between 0.35 and 0.4645 values between 0.46 and 0.5625 values between 0.56 and 0.6615 values between 0.66 and 0.7611 values between 0.76 and 0.868 values between 0.86 and 0.964 values between 0.96 and 1.074 values between 1.07 and 1.171 values between 1.17 and 1.272 values between 1.27 and 1.374 values between 1.37 and 1.473 values between 1.47 and 1.571 values between 1.88 and 1.982 values between 2.39 and 2.49offnegativepositiveoften

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. 主场
  2. 行った
  3. 的那种
  4. build
  5. .build
  6. 행
  7. 星球
  8. pios
  9. builder
  10. ::_

Suppresses

  1. amac
  2. atem
  3. �数
  4. stantiate
  5. 日起
  6. etration
  7. "><?=
  8. cog
  9. quez
  10. quet