Neuron Explorer

Qwen3-0.6B

numbers and dates with stray punctuation

A neuron in layer 0 of Qwen3-0.6B. Its strongest snippets are mixed: they peak on different tokens. It is featured, so it has steering outputs.

What this neuron does

This neuron reacts to small punctuation pieces in numbers and dates, like the dot in "2.M" or the dash in "1910-Moved". It is a weak signal, and turning it up gave text that did not show any clear theme.

Written by Claude from this neuron's activations and steering outputs. Confidence: low. 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

    2 students learn to tell and write time from analog and digital clocks to the nearest five minutes, using a.m. and p.m.⏎Common Core: 2.MD.7⏎

  2. #2 peak

    of Pharmacy terminated.⏎1905-Theological Seminary organized in Waco.⏎1907-Separated from Baylor University.⏎1910-Moved to Fort Worth.⏎1918-College of Dentistry organized in

  3. #3 peak

    along with other associations, sent representatives in 1848 to create the Baptist State Association, which later became the Baptist State Convention.⏎1845-Baylor chartered on February 1 by the Republic of Texas.⏎184

  4. #4 peak

    .⏎1996 -The College became a part of the Texas A&M System on September 1, 1996.⏎1919-Baylor Hospital organized in Dallas – now Baylor University Medical Center.⏎1919

  5. #5 peak

    _304 Measurement System Inch Number of Items 1 Part Number 190.0400.0400.025.SS UNSPSC Code 30260000 See more

  6. #6 peak

    80E.2s in 1919, joining seven Nieuport 83.E.2s that they had purchased in 1916.⏎In

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 have been thinking about how to get a good score in a game. And I have been thinking about how to get a good score in a game. I have been thinking about how to get a good score in a game. And 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

    North American continent.⏎Rising prominently at 5,895 m (19, 341 ft) in north-eastern Tanzania, the snow-capped

  2. #2 peak

    72,000 km/hour (45,000 miles/hour) when it hit the atmosphere. As an asteroid

  3. #3 peak

    m (18,510 ft) on its western summit and 5,621

  4. #4 peak

    Curtiss and purchase a plane for $7,000 ($157,000 in today’s money). While waiting for it to be built,

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. L4:2598 suppresses
  2. L8:592 weakly suppresses
  3. L6:2063 weakly excites
  4. L7:1654 weakly excites 'memory' in brain and ageing text
  5. L26:385 weakly suppresses
  6. L6:1240 weakly excites
  7. L2:1039 weakly excites
  8. L19:0 weakly excites
  9. L5:365 weakly suppresses
  10. L6:697 weakly excites
  11. L6:58 weakly suppresses
  12. L3:2392 weakly excites
  13. L4:779 weakly excites
  14. L16:1218 weakly suppresses
  15. L7:2614 weakly excites
  16. L20:628 weakly excites
  17. L7:2421 weakly excites
  18. L22:2385 weakly excites
  19. L25:112 weakly suppresses
  20. L6:182 weakly excites
  21. L3:732 weakly excites
  22. L22:1316 weakly excites
  23. L9:906 weakly excites
  24. L10:780 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 -1.30 and -1.261 values between -1.26 and -1.211 values between -1.21 and -1.171 values between -0.99 and -0.941 values between -0.94 and -0.901 values between -0.85 and -0.811 values between -0.54 and -0.501 values between -0.50 and -0.451 values between -0.45 and -0.412 values between -0.41 and -0.364 values between -0.36 and -0.321 values between -0.32 and -0.276 values between -0.27 and -0.2311 values between -0.23 and -0.1852 values between -0.18 and -0.14169 values between -0.14 and -0.09465 values between -0.09 and -0.051110 values between -0.05 and -0.001224 values between -0.00 and 0.04402 values between 0.04 and 0.09207 values between 0.09 and 0.1358 values between 0.13 and 0.1822 values between 0.18 and 0.2222 values between 0.22 and 0.2732 values between 0.27 and 0.3142 values between 0.31 and 0.3648 values between 0.36 and 0.4020 values between 0.40 and 0.452 values between 0.45 and 0.49offnegativepositiveoften

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. amus
  2. ró
  3. elow
  4. ATA
  5. rush
  6. !
  7. !
  8. ingroup
  9. -pro
  10. ở

Suppresses

  1. 杀手
  2. 利用率
  3. 计算器
  4. 打击
  5. tester
  6. 鸡肉
  7. thon
  8. setSupportActionBar
  9. banker
  10. eteria