Neuron Explorer

Qwen3-0.6B

"with" after combine, mix or react verbs

A neuron in layer 11 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

Fires on the word "with" when it follows a verb of joining or reacting, like "combined with", "mixed with" or "reacts with". Turning it up made the model drift into rambling, unrelated text.

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

    through a wire to create an electric current. The electrons complete the detour to meet up with the ions in the final segment of the fuel cell where they combine with oxygen to create water or carbon dioxide. The energy produced is clean, with water

  2. #2 peak

    BODY.⏎Now upon death, the intelligence of the person, automatically within seconds enters the body of another child that is 11 years. It then combines itself with the intelligence that child had from birth. Example, if a man became a doctor

  3. #3 peak

    is mined, melted, and split into hydrogen and oxygen by electrolysis. The lakes' hydrocarbons are pumped from the lakes into stations where they are mixed with the oxygen derived from the ice—that is to say, burned—resulting in

  4. #4 peak

    decided that the licensing provisions were constitutional and reversed the appellate court. The Jehovah’s Witnesses from Alabama appealed to the U.S. Supreme Court, which consolidated the case with two other licensing fee cases involving Witnesses.⏎Court originally upheld license requirement⏎Justice Stanley

  5. #5 peak

    there is blooming vegetation bordering the treated field. EPA is prohibiting tank mixing of sulfoxaflor with pesticides that have shown evidence of synergistic activity with sulfoxaflor. The product label directs applicators to more information and

  6. #6 peak

    “Hydrogen is quite a versatile thing,” Constantine says. “So if we’re looking at breweries in particular, they can use that hydrogen to blend with natural gas for heating or they can also convert it back into electricity for renewables if

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: "A single sentence with a single idea, and then a few sentences are distributed across the sentence structure, which is the primary unit. In the case of a single sentence, there is a single idea, and the sentence structure is typically a

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

    0" long, and 8" wide⏎Attributes: Good for Containers⏎When to Sow Outside: RECOMMENDED. 4 to 6 weeks before your average last frost date, when soil temperature is at least 40°F

  2. #2 peak

    How to Plant Marigolds⏎- If you wish to start growing marigolds in spring, sow them indoors, particularly six to eight weeks before your area’s last frost date. If you are fine with later blooms, sow

  3. #3 peak

    .⏎- Use voice commands⏎- Nowadays, AI is becoming more intelligent. We can use voice commands instead of typing on the screen.⏎- Avoid sharing your devices with others⏎- If unnecessary, do not share your smartphone or other devices with others

  4. #4 peak

    014⏎One approach if giving criticism is to put it inside a sandwich- open with aspects you praise or agree with, offer critical statements, and close with a positive. Maybe the best advice is to comment in the style and mood that

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. L6:800 excites
  2. L9:1252 suppresses
  3. L10:2213 suppresses 'to' or 'into' showing movement between two things
  4. L9:2267 excites 'with' after exchange nouns
  5. L9:1820 excites
  6. L8:150 excites
  7. L7:1949 excites
  8. L7:2161 excites
  9. L2:2260 excites 'a' after 'in' or 'such'
  10. L6:1828 suppresses
  11. L8:2999 suppresses
  12. L6:364 excites
  13. L6:1084 suppresses
  14. L9:878 suppresses
  15. L10:515 weakly suppresses
  16. L3:2502 weakly suppresses
  17. L8:1083 weakly excites "x" or "by" between two dimensions
  18. L5:910 weakly excites
  19. L9:1144 weakly excites
  20. L8:809 weakly excites
  21. L8:2596 weakly excites
  22. L9:1908 weakly excites
  23. L4:502 weakly excites
  24. L10:1359 weakly suppresses

Feeds later layers

  1. L15:2247 excites
  2. L14:2830 excites
  3. L14:2100 excites
  4. L14:1654 excites
  5. L12:2739 excites
  6. L19:2698 excites
  7. L12:258 excites
  8. L15:2205 excites
  9. L17:2769 excites
  10. L15:2522 excites
  11. L12:1394 excites
  12. L16:1469 excites
  13. L15:3021 excites
  14. L13:716 excites
  15. L13:2754 excites
  16. L14:2320 excites
  17. L12:1454 excites
  18. L12:2930 excites "and" joining contrasting pairs
  19. L19:1572 excites
  20. L15:1037 excites
  21. L18:2256 excites
  22. L12:1907 excites
  23. L12:1544 suppresses
  24. L15:361 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.

3 values between -1.09 and -0.992 values between -0.99 and -0.882 values between -0.88 and -0.783 values between -0.78 and -0.6811 values between -0.68 and -0.5719 values between -0.57 and -0.4751 values between -0.47 and -0.37167 values between -0.37 and -0.26433 values between -0.26 and -0.16932 values between -0.16 and -0.061508 values between -0.06 and 0.05513 values between 0.05 and 0.15155 values between 0.15 and 0.2542 values between 0.25 and 0.3520 values between 0.35 and 0.469 values between 0.46 and 0.5615 values between 0.56 and 0.665 values between 0.66 and 0.773 values between 0.77 and 0.872 values between 0.87 and 0.973 values between 0.97 and 1.082 values between 1.08 and 1.181 values between 1.28 and 1.391 values between 1.39 and 1.491 values between 1.49 and 1.591 values between 1.59 and 1.691 values between 1.69 and 1.802 values between 1.90 and 2.001 values between 2.93 and 3.03offnegativepositiveoften

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. sake
  2. оборот
  3. Relationship
  4. 一起
  5. entifier
  6. работу
  7. enal
  8. 搭配
  9. gorit
  10. 同事

Suppresses

  1. zano
  2. Fra
  3. Chambers
  4. quo
  5. pons
  6. あげ
  7. Hawkins
  8. -pill
  9. hog
  10. /qu