Neuron Explorer

Qwen3-0.6B

the word 'system' or 'systems'

A neuron in layer 0 of Qwen3-0.6B. Its strongest snippets are fairly consistent: several of them peak on the same token. It is featured, so it has steering outputs.

What this neuron does

This neuron lights up on the word "system" or "systems" when it means a set of connected parts, like food systems or herbal medicine systems. When it was turned up, the text drifted into rambling, repeated phrases with no 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

    ’s most important witnesses are the firefighters that first arrive on the scene. The investigator will ask firefighters for details such as the color of the smoke, damaged sprinkler systems, condition of doors and windows, and general strange behavior of the flames. After

  2. #2 peak

    and treating the root cause of health problems as opposed to treating symptoms.⏎Herbal medicine is the use of plant remedies in the treatment of disease. As the oldest system of medicine known, it is employed by more than 80 per cent of

  3. #3 peak

    gender, and class as they effect the personal problems of individual families While keeping an explicit focus on clinical work, this volume highlights power differentials in families and larger systems⏎What do feminist think about the nuclear family?⏎2 Feminists argue that in

  4. #4 peak

    item, piece of equipment, or product system, whether acquired commercially off the shelf, modified, or customized, that is used

  5. #5 peak

    for a kidney, and<|endoftext|>Cornell Cooperative Extension Albany County is unique among local organizations in offering a continuum of education and resources that enable it to address a food systems issues in a holistic way. By working at a food-system level, we can

  6. #6 peak

    for our country.” If this is the case, why not turn this awareness into action this Earth Day?⏎What changes can you make in the areas of…⏎Food systems influence many environmental factors. How is it sourced, processed, and distributed? Did

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: > I was thinking about lately, I was thinking about lately. < So I was thinking about lately. < So, I was thinking about lately. < So, I was thinking about lately. < So, I

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

    of the natural world and their place in it, encouraging a deeper understanding of oneself and the surrounding environment.⏎Q2: How does Mary<|endoftext|>Unified Modeling Language⏎<language> (UML) A non-proprietary, third generation modelling

  2. #2 peak

    , to encourage bushiness. However, if you want to enjoy the beautiful flowers, allow some stems to grow, so they can form buds and blooms.<|endoftext|>< Browse to Previous Essay | Browse to Next Essay >⏎Seattle unveils steam-powered fire

  3. #3 peak

    therapy may be far more effective at killing all cancerous cells than removing the entire breast.⏎Lead researcher Dr E Shelley Hwang, of the Duke Cancer Institute in <a class=”StrictlyAutoTagAnchor” title=”View all articles about North

  4. #4 peak

    national sample. Lifetime use rates for American Indian students were higher than those of the national sample for all substance measures except tranquilizers and amphetamines (all P <.05). The greatest discrepancy was observed for 8th grade students:

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

None: this is the first layer.

Feeds later layers

  1. L19:1206 suppresses
  2. L4:1075 suppresses
  3. L18:805 suppresses
  4. L3:566 weakly suppresses
  5. L4:1850 weakly excites
  6. L1:1855 weakly suppresses
  7. L14:2736 weakly excites
  8. L27:862 weakly excites
  9. L15:320 weakly suppresses
  10. L18:1721 weakly excites
  11. L13:1992 weakly excites
  12. L8:545 weakly suppresses
  13. L22:1135 weakly excites
  14. L4:203 weakly suppresses
  15. L3:2689 weakly excites
  16. L13:2038 weakly excites
  17. L11:1953 weakly excites
  18. L22:1948 weakly excites
  19. L11:1943 weakly excites
  20. L1:297 weakly excites
  21. L5:133 weakly excites
  22. L2:1682 weakly suppresses
  23. L21:2744 weakly excites
  24. L7:601 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.

1 values between -0.58 and -0.552 values between -0.55 and -0.511 values between -0.39 and -0.362 values between -0.36 and -0.332 values between -0.33 and -0.302 values between -0.30 and -0.273 values between -0.27 and -0.245 values between -0.24 and -0.2111 values between -0.21 and -0.1825 values between -0.18 and -0.1552 values between -0.15 and -0.12123 values between -0.12 and -0.09299 values between -0.09 and -0.05668 values between -0.05 and -0.021556 values between -0.02 and 0.01658 values between 0.01 and 0.04278 values between 0.04 and 0.0791 values between 0.07 and 0.1054 values between 0.10 and 0.1326 values between 0.13 and 0.1614 values between 0.16 and 0.194 values between 0.19 and 0.2211 values between 0.22 and 0.256 values between 0.25 and 0.285 values between 0.28 and 0.313 values between 0.31 and 0.342 values between 0.34 and 0.381 values between 0.38 and 0.411 values between 0.47 and 0.501 values between 0.53 and 0.561 values between 0.62 and 0.65offnegativepositiveoften

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. smuggling
  2. 貂
  3. PKK
  4. 粹
  5. oca
  6. doGet
  7. GetLastError
  8. smugg
  9. uiten
  10. Arabic

Suppresses

  1. chron
  2. ее
  3. ==⏎
  4. IENT
  5. ies
  6. >(⏎
  7. itches
  8. ptime
  9. емся
  10. IBLE