Neuron Explorer

Qwen3-0.6B

"and" or "or" in a pair or list

A neuron in layer 15 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 the word "and" (or "or") when it joins two or three things of the same kind, like "silver, tin, and copper" or "men and women". Turning it up made the model repeat phrases, and turning it down gave a more general, flowing style.

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

    two ends on the same continuum?⏎Homer, Plotinus, and Plato have all posited answers to this question, whether directly - in the case of Plato and Plotinius - or through story - in the case of Homer.⏎Plotinus

  2. #2 peak

    "silver" are actually a mixture of silver, tin, copper, and mercury, and are more properly called amalgam fillings. The silver, tin, and copper are provided in a powder form by various manufacturers, with silver being the major

  3. #3 peak

    . In the late nineteenth century a Scottish Mill designed and produced four cloths, one for each of the world’s most distinguished universities. While Harvard, Yale, and Cambridge all had their respective cloths lost to history, for reasons unknown Oxford endured

  4. #4 peak

    the FCAT?⏎The FCAT includes multiple-choice, gridded-response (fill in the blanks) and performance tasks (such as essays). The multiple-choice and gridded-response questions are machine scored. Each performance task test is scored by

  5. #5 peak

    up.⏎- Wear clothing that will allow the skin to breathe, so you don’t become overheated.⏎- A hat helps to hold the heat in during winter and protects the eyes in the summer.⏎The way you hold the body while running is

  6. #6 peak

    or more for women. The study found that of the almost 50,000 people surveyed across China, 55.6% of men and 15% of women were current drinkers, having had at least one drink

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: how to use the single, single, and the single, in the single, single, and in, on the single, without, on the single, the single, and all, in the single, and, by, on the single

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

    , movies and other licensed digital information resulting in loss of potential revenue. Examples include, attack on the national infrastructure such as electricity grid, electricity and water, delivery system, air traffic controls that depend on communications and network. This

  2. #2 peak

    negative. You blood groups help doctors know many things, like what kind of blood you’ll need in a blood transfusion or an organ donation, and if you or your baby will be in danger if you get pregnant.⏎So how does your body

  3. #3 peak

    in a discontinuity analysis on observations near the borders between anglophone and francophone countries.⏎- Colonial rule⏎ASJC Scopus subject areas⏎- Economics and Econometrics⏎- Organizational Behavior and Human Resource Management<|endoftext|>- About ASTI

  4. #4 peak

    105° Fahrenheit).⏎How can I prepare?⏎- Listen to local weather forecasts and stay aware of upcoming temperature changes.⏎- Be aware of both the temperature and the heat index. The heat index is the temperature the body feels when the effects

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. L6:427 suppresses
  2. L12:2058 excites
  3. L14:2202 excites
  4. L14:801 excites 'and' or 'or' joining list items
  5. L12:1786 excites
  6. L7:1223 suppresses "and" joining two decades or times
  7. L12:51 excites
  8. L14:2865 suppresses
  9. L9:92 suppresses "and" or "or" joining two items
  10. L2:1000 excites
  11. L13:1554 excites
  12. L11:406 weakly suppresses
  13. L12:1336 weakly suppresses
  14. L10:2046 weakly suppresses
  15. L9:738 weakly suppresses
  16. L1:1250 weakly suppresses
  17. L9:2010 weakly suppresses
  18. L9:77 weakly excites commas in citations and name lists
  19. L4:57 weakly suppresses
  20. L13:2931 weakly excites
  21. L13:131 weakly excites 'and' joining parallel phrases
  22. L5:880 weakly excites
  23. L7:836 weakly excites
  24. L12:2322 weakly suppresses citation and catalogue metadata text

Feeds later layers

  1. L16:1937 excites 'and' joining two time or quantity ranges
  2. L16:2454 weakly suppresses
  3. L18:2354 weakly excites
  4. L16:2577 weakly excites
  5. L16:2940 weakly excites commas inside lists and names
  6. L16:1805 weakly excites
  7. L16:2145 weakly excites
  8. L16:2568 weakly suppresses
  9. L16:1692 weakly suppresses
  10. L17:2623 weakly excites
  11. L16:279 weakly suppresses
  12. L20:364 weakly excites
  13. L18:1975 weakly excites
  14. L18:2385 weakly suppresses
  15. L17:1610 weakly suppresses
  16. L20:1009 weakly excites
  17. L23:1275 weakly suppresses
  18. L20:1458 weakly suppresses
  19. L18:2037 weakly suppresses
  20. L25:2927 weakly excites
  21. L16:2296 weakly excites
  22. L25:1758 weakly excites
  23. L20:901 weakly excites
  24. L17:1022 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 -3.25 and -3.061 values between -3.06 and -2.864 values between -2.28 and -2.091 values between -2.09 and -1.904 values between -1.90 and -1.706 values between -1.70 and -1.515 values between -1.51 and -1.322 values between -1.32 and -1.137 values between -1.13 and -0.9312 values between -0.93 and -0.7416 values between -0.74 and -0.5555 values between -0.55 and -0.35437 values between -0.35 and -0.162224 values between -0.16 and 0.03820 values between 0.03 and 0.23138 values between 0.23 and 0.4260 values between 0.42 and 0.6127 values between 0.61 and 0.8121 values between 0.81 and 1.0018 values between 1.00 and 1.196 values between 1.19 and 1.387 values between 1.38 and 1.586 values between 1.58 and 1.774 values between 1.77 and 1.965 values between 1.96 and 2.163 values between 2.16 and 2.354 values between 2.35 and 2.542 values between 2.54 and 2.742 values between 2.74 and 2.932 values between 2.93 and 3.123 values between 3.12 and 3.322 values between 3.51 and 3.702 values between 3.89 and 4.091 values between 4.28 and 4.47offnegativepositiveoften

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. idad
  2. itate
  3. itating
  4. We
  5. osci
  6. we
  7. ista
  8. razy
  9. itated
  10. otle

Suppresses

  1. 骝
  2. 的支持
  3. FE
  4. 告诉她
  5. Registration
  6. igner
  7. 另一边
  8. aar
  9. 叮
  10. Qualität