Neuron Explorer

Qwen3-0.6B

the last zero in large numbers

A neuron in layer 3 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 a zero inside a big number, like the 0 in 1500, 2000 or 800. When it was turned up, the model's replies drifted into rambling text with no clear link to numbers.

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

    and brain.⏎What Vitamins Should heavy drinkers take?⏎- Include 250mg Vitamin C, 150mg magnesium, 1500mg calcium and 500 mg niacin from dietary sources each day.

  2. #2 peak

    ] stamp, this stamp is supposed to certify the produce is 90% organic non GMO. So why no seeds?⏎viewed 15008 times⏎People have been modifying plants for thousands of years. This is called

  3. #3 peak

    War II.⏎Here are some Fw 190’s specs at a glance:⏎- Maximum speed-408 mph⏎- Range-500 miles⏎- Service

  4. #4 peak

    1 / √2⏎E) 2 √2⏎What is the centripetal force acting on a car of mass 2000 Kg turning around a circular bend of radius 50 meters at the constant speed

  5. #5 peak

    so they are very clean. You can also disinfect them with a mild bleach solution.⏎- yoghurt containers – the small 8 ounce (250ml) single serving sized ones are

  6. #6 peak

    DPI laser mice are very popular, but some high-end models are capable of 1000, 1600 or even 2000 DPI tracking speeds.⏎The Personal System/2 or PS/2 was the designation

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 am not a fan with this. I have been on the beach of this thing. So, the first thing is with my head. I am not a person. I'm a person. Okay, this is about me. I have

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

    There are many calculators out there which can help you determine your energy expenditure. One that seems<|endoftext|>Italian scientists have found another image on the back of the Shroud of Turin, reported The Guardian yesterday. The scientists from the University of Pad

  2. #2 peak

    the result of an intense source of energy as the body passed through the cloth, perhaps at the moment of resurrection,” reported The Guardian.⏎Believed to be shroud of Jesus⏎The man, whose image appears on the shroud, seems to

  3. #3 peak

    ies, even if abandoned, in ruined condition or transformed for other use, also served as symbolic mezuzahs to mark towns, villages, cities, and even countries where Jews once

  4. #4 peak

    Stevenson House was the centre of his world and his mind was nourished by ceaseless reading as well as the stories told by Cummie of ghosts, ghouls, Scottish history, and the Bible. He also travelled with his father around

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. L0:1301 weakly suppresses
  2. L0:1224 weakly suppresses
  3. L0:178 weakly suppresses
  4. L1:2659 weakly excites
  5. L2:1795 weakly suppresses
  6. L2:79 weakly excites
  7. L2:1377 weakly suppresses
  8. L0:2825 weakly excites
  9. L2:1657 weakly excites
  10. L2:237 weakly excites
  11. L1:2544 weakly suppresses
  12. L1:2411 weakly excites
  13. L0:2352 weakly excites
  14. L2:2716 weakly suppresses
  15. L1:469 weakly excites
  16. L2:462 weakly excites
  17. L2:366 weakly suppresses
  18. L0:2721 weakly excites
  19. L0:252 weakly excites
  20. L1:2057 weakly excites
  21. L0:1646 weakly suppresses
  22. L2:49 weakly suppresses
  23. L0:16 weakly excites
  24. L2:1268 weakly excites 'ur' and 'ud' after a capital H

Feeds later layers

  1. L4:2836 weakly excites
  2. L16:1962 weakly suppresses
  3. L7:1218 weakly suppresses
  4. L25:2025 weakly excites
  5. L6:2516 weakly suppresses
  6. L8:3000 weakly excites
  7. L14:2085 weakly excites
  8. L11:1552 weakly excites
  9. L18:2666 weakly suppresses
  10. L10:672 weakly suppresses
  11. L9:750 weakly excites
  12. L18:2908 weakly excites
  13. L23:324 weakly suppresses
  14. L7:2017 weakly excites
  15. L15:2715 weakly excites
  16. L10:95 weakly suppresses
  17. L27:961 weakly suppresses
  18. L11:1674 weakly excites
  19. L20:631 weakly suppresses
  20. L10:476 weakly suppresses
  21. L7:514 weakly suppresses 'always' or 'often' after 'is' or 'there is'
  22. L4:2550 weakly suppresses
  23. L7:1169 weakly excites
  24. L7:2232 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.

2 values between -1.11 and -0.911 values between -0.91 and -0.701 values between -0.70 and -0.5037 values between -0.50 and -0.29768 values between -0.29 and -0.082648 values between -0.08 and 0.12411 values between 0.12 and 0.3320 values between 0.33 and 0.542 values between 0.54 and 0.741 values between 0.95 and 1.151 values between 1.15 and 1.362 values between 1.57 and 1.771 values between 1.77 and 1.981 values between 2.19 and 2.391 values between 2.80 and 3.011 values between 3.01 and 3.221 values between 3.22 and 3.421 values between 3.63 and 3.831 values between 4.04 and 4.251 values between 4.25 and 4.451 values between 5.07 and 5.281 values between 5.69 and 5.901 values between 5.90 and 6.101 values between 6.10 and 6.312 values between 6.93 and 7.13offnegativepositiveoften

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. 脲
  2. Raqqa
  3. レーション
  4. IRM
  5. plies
  6. ."</
  7. leitung
  8. berman
  9. sWith
  10. ta

Suppresses

  1. Voor
  2. ius
  3. 凡
  4. Stat
  5. Accept
  6. prom
  7. ény
  8. [m
  9. nost
  10. eway