Neuron Explorer

Qwen3-0.6B

the word 'of' in 'middle of'

A neuron in layer 0 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 lights up on the word 'of' when it comes right after 'middle', as in 'in the middle of the road'. Turning it up made the model's text drift into unrelated topics, so the steering results don't add much.

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

    . The audience shock-factor came from the garment that Bella Hadid wore, which was a spray-on dress. Hadid walked slowly onto a platform in the middle of the runway. Here she was sprayed by two technicians with a white material. With

  2. #2 peak

    's where Rob Johnstone chooses to spend his spare time.⏎As the sun begins to set, the astronomer drives out to an abandoned airstrip in the middle of the desert and sets up his complicated telescope as the remnants of the day's light

  3. #3 peak

    RAND's Robert Lempert studies climate change risk management. In October, his professional and personal lives collided when he was ordered to evacuate his home in the middle of the night to escape a spreading wildfire. Lempert's experience emphasized the challenges

  4. #4 peak

    sun. Wear sunglasses, a hat and a shirt, especially in the middle of the day (even on overcast days).⏎Eat right. Include at least five

  5. #5 peak

    deer behavior on the roads, sharing stories of deer not only darting out into the road but jumping right into the side of the car or stopped in the middle of the road, "caught in the headlights."⏎This video (thanks to National Forest

  6. #6 peak

    difference.⏎When does a child’s apology matter to parents?⏎<|endoftext|>Deep in a North Carolina marsh, a lone swamp sparrow sits on his perch in the middle of the water. He’s singing his usual song. But he’s also aggressively fl

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 think that the problem is to find the number of the number of the number of the number of the number of the number of the number of the number of the number of the number of the number of the number of the number of the

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

    Harvard T.H. Chan School of Public Health, Sept. 2012, https://www.hsph.harvard.edu/nutritionsource/vitamins/.⏎- Betta Health, VIC Government

  2. #2 peak

    0 through Summer 2011. The departments listed here offer sustainability course and programs. View the current complete course listings here.⏎A “sustainability-focused” course concentrates on the concept of sustainability, including its social, economic, and environmental

  3. #3 peak

    Indo-Aryan language—the Tamils speak the Dravidian language of Tamil. Again, while more than 90 percent of the Sinhalese are Buddhists, both Sri

  4. #4 peak

    -an-acre program. It is not uncommon for loggers, oil and gas companies, cattle ranchers, and miners to illegally extract resources from so-called “protected” areas.⏎Traditional forest communities are often the best stewards of the land because their

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. L22:271 strongly suppresses
  2. L21:625 strongly excites
  3. L18:853 suppresses
  4. L22:1055 excites
  5. L2:761 excites
  6. L23:50 suppresses
  7. L23:474 suppresses
  8. L24:124 suppresses dollar sign before a money amount
  9. L21:1781 excites
  10. L27:1687 suppresses
  11. L21:3030 suppresses
  12. L21:2021 suppresses the 4 in years 184x, 164x
  13. L3:459 suppresses
  14. L20:854 excites
  15. L24:204 suppresses
  16. L25:2877 suppresses
  17. L4:524 excites
  18. L21:2500 excites
  19. L25:231 excites
  20. L26:319 suppresses
  21. L25:36 suppresses
  22. L23:1351 suppresses
  23. L20:2688 excites
  24. L4:216 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.

4 values between -0.75 and -0.711 values between -0.71 and -0.681 values between -0.68 and -0.641 values between -0.58 and -0.542 values between -0.54 and -0.511 values between -0.51 and -0.481 values between -0.48 and -0.442 values between -0.41 and -0.385 values between -0.34 and -0.316 values between -0.31 and -0.2810 values between -0.28 and -0.2421 values between -0.24 and -0.2129 values between -0.21 and -0.1738 values between -0.17 and -0.1471 values between -0.14 and -0.11102 values between -0.11 and -0.07229 values between -0.07 and -0.04616 values between -0.04 and -0.011285 values between -0.01 and 0.03638 values between 0.03 and 0.06351 values between 0.06 and 0.10220 values between 0.10 and 0.13116 values between 0.13 and 0.1665 values between 0.16 and 0.2045 values between 0.20 and 0.2315 values between 0.23 and 0.2611 values between 0.26 and 0.309 values between 0.30 and 0.334 values between 0.33 and 0.363 values between 0.36 and 0.402 values between 0.40 and 0.432 values between 0.50 and 0.531 values between 0.53 and 0.571 values between 0.57 and 0.60offnegativepositiveoften

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. gether
  2. ly
  3. theless
  4. ン
  5. hip
  6. ful
  7. plier
  8. LY
  9. warts
  10. ся

Suppresses

  1. .persistent
  2. ×⏎⏎
  3. hình
  4. 形
  5. ares
  6. pressions
  7. 了
  8. gráfico
  9. PRESSION
  10. unique