Neuron Explorer

Qwen3-0.6B

colon in clock times

A neuron in layer 14 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 colon that separates hours from minutes in a time stamp, like '08:57' or '11:07 am'. It is a clock-time detector. Turning it up gave text that did not obviously relate to times, so the writing effect is unclear.

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

    Jan 09, 2017 08:57 AM EST⏎MIT Professor Is Building Bio-Inspired Robots To Save Lives⏎Dis

  2. #2 peak

    turn directions, UI, User interface, video chat, Virtual reality, Virtual reality glasses⏎January 18, 2012 • 11:07 am

  3. #3 peak

    But that merit might be offset if the Chinese consume relentlessly like<|endoftext|>Lev Bely, 07.05.2012 21:56⏎There are still butterflies in Mumbai albeit they're increasingly getting lack of

  4. #4 peak

    ||World match.|⏎|Next Article:||Ukraine's Orange Revolution.|<|endoftext|>posted on November 17, 2014 23:23⏎Scientists from the WCS (Wildlife Conservation Society), NASA, and

  5. #5 peak

    SeaWeb/FIS)⏎Seafood producers tackle ocean acidification at annual Summit⏎Monday, February 21, 2011, 23:50 (GMT + 9)⏎From 31 January to 2

  6. #6 peak

    subject is much smaller than a human. Essential oils are highly concentrated and should<|endoftext|>Monday, October 5, 1998 Published at 17:47 GMT

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 can't, but I can. I can. I can. I can. I can. I can. I can. I can. I can. I can. I can. I can. I can. I can. 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

    regulating your intake and staying healthy, both physically and mentally.<|endoftext|>Tech moves fast! Stay ahead of the curve with Techopedia!⏎Join nearly 200,000 subscribers who receive actionable tech insights from Techopedia.⏎A

  2. #2 peak

    oes, tsunamis, global climate, and weather.⏎Based on your research, do the following:⏎Support your statements with examples. Use a minimum of six reliable references, two of which should be peer-reviewed articles.⏎Write a 7

  3. #3 peak

    July 10, the fire had damaged or destroyed nearly 130 buildings, injured over 20 people, and cut the lives of 19 Granite Mountain Hotshots woefully short.⏎Sparked by Lightning, Fueled

  4. #4 peak

    established in 1996 to recognize, honor, and promote the outstanding contributions women make to the scientific and technological communities that improve society. Each year, five women are selected from around the world to receive this honor, and Trayanova

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. L7:1524 suppresses
  2. L10:993 excites
  3. L4:87 suppresses
  4. L3:203 weakly suppresses
  5. L10:50 weakly excites
  6. L7:224 weakly suppresses
  7. L7:830 weakly suppresses
  8. L8:582 weakly suppresses
  9. L4:538 weakly suppresses
  10. L12:408 weakly suppresses
  11. L13:408 weakly suppresses
  12. L4:201 weakly excites
  13. L8:758 weakly excites
  14. L9:183 weakly suppresses
  15. L13:2861 weakly suppresses
  16. L11:93 weakly excites
  17. L13:1215 weakly suppresses
  18. L7:723 weakly suppresses
  19. L13:1299 weakly suppresses
  20. L13:1297 weakly suppresses
  21. L13:460 weakly suppresses
  22. L13:2705 weakly excites
  23. L0:2820 weakly suppresses
  24. L12:953 weakly excites

Feeds later layers

  1. L16:314 excites
  2. L17:2532 suppresses
  3. L16:122 excites
  4. L18:467 weakly excites
  5. L18:411 weakly suppresses
  6. L19:785 weakly suppresses
  7. L21:1972 weakly suppresses
  8. L16:2787 weakly suppresses
  9. L24:1032 weakly excites
  10. L21:299 weakly excites
  11. L15:1080 weakly suppresses
  12. L27:1321 weakly suppresses
  13. L26:682 weakly suppresses
  14. L21:110 weakly suppresses
  15. L17:248 weakly excites
  16. L24:2630 weakly excites
  17. L25:2001 weakly suppresses
  18. L26:1246 weakly suppresses
  19. L19:42 weakly excites
  20. L17:167 weakly excites
  21. L25:182 weakly suppresses
  22. L24:272 weakly suppresses
  23. L16:1212 weakly excites
  24. L26:1451 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 -6.75 and -6.512 values between -6.27 and -6.031 values between -5.78 and -5.541 values between -4.81 and -4.571 values between -4.57 and -4.332 values between -4.33 and -4.083 values between -4.08 and -3.845 values between -3.84 and -3.601 values between -3.60 and -3.362 values between -3.36 and -3.115 values between -3.11 and -2.874 values between -2.87 and -2.631 values between -2.63 and -2.383 values between -2.38 and -2.148 values between -2.14 and -1.907 values between -1.90 and -1.665 values between -1.66 and -1.4111 values between -1.41 and -1.1718 values between -1.17 and -0.9311 values between -0.93 and -0.6934 values between -0.69 and -0.44178 values between -0.44 and -0.202245 values between -0.20 and 0.041221 values between 0.04 and 0.29100 values between 0.29 and 0.5314 values between 0.53 and 0.779 values between 0.77 and 1.025 values between 1.02 and 1.265 values between 1.26 and 1.501 values between 1.50 and 1.741 values between 1.74 and 1.991 values between 1.99 and 2.231 values between 2.47 and 2.711 values between 2.71 and 2.96offnegativepositiveoften

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. leurs
  2. SON
  3. son
  4. 落实
  5. 突破
  6. s
  7. 適
  8. 依靠
  9. 加重
  10. H

Suppresses

  1. iation
  2. /reference
  3. iative
  4. iated
  5. iator
  6. attles
  7. 骸
  8. iate
  9. bruary
  10. Pompeo