Neuron Explorer

Qwen3-0.6B

comma inside dates and years

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 fires on the comma inside a citation or date, such as the comma after a year in "(2014, July)" or "(1990, 1992)". Its strongest negative cases are the word "to" in number ranges like "from 1792 to today" or "1210 to 1274". Turning it up did not give a clear theme in the steered text.

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

    M., Chan, K. S.-C., Chan, D. W.-O., & Chung, K. K.-H. (2014, July). Effectiveness of a tiered intervention model in Chinese: A growth model

  2. #2 peak

    the module.⏎See the Discussion Board Guidelines and Rubrics⏎Download Discussion Board Guidelines and Rubrics.⏎Reh, F. J. (2018, November 01). Here Is a Look at the Role and Responsibilities of Today

  3. #3 peak

    the extent and patterns of seasonality in FC levels in coastal waters. Water quality monitoring data for shellfish management collected during El Niño (1990, 1992, 1997), and La

  4. #4 peak

    of having an opportunity to prosper in life.⏎The following two case studies are based on the book The Life Span Human Development for Helping Professionals (2006, 2nd ed. ) by Patriia C.⏎John Winthrops City

  5. #5 peak

    Probes of Bulk and Surface Compositions of Size-Resolved Sea Spray Aerosol Particles,” Phys. Chem. Chem. Phys., 2013, 15, 6206-6214.⏎Atomic

  6. #6 peak

    ideal supplement to high school and college-level Holocaust studies, and adult interfaith discussion groups.⏎A Time to Remember⏎USA, 1992, 52 minutes, color⏎Produced by Sy Rotter and Andrzej

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: 10, 20, 30, 40, 50, 60, 70, 80, 90, 100, 110, 1

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

    House? Learn about this and other presidential "firsts."⏎Read more >⏎History of the White House and Washington, D.C. from 1792 to today.⏎Who was the first U.S. president to be photographed?⏎Grades

  2. #2 peak

    ization.⏎HG is a more severe form of nausea and vomiting of pregnancy (NVP), also called morning sickness. Morning sickness affects anywhere between 50% to 90% of pregnant women. HG is estimated to occur in 0

  3. #3 peak

    9. Coutances houses a well-known botanical garden and an art museum.⏎Coutances Cathedral is a Gothic Roman Catholic cathedral constructed from 1210 to 1274 in the town of Coutances, Normandy, France

  4. #4 peak

    a Temple choir and a Board of Relief established to care for Jewish burials and illnesses. Rabbi Jacob Block served B’nai Israel from 1872 to 1880.⏎Disagreements about religious observances and liturgy

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. L2:1710 weakly suppresses
  2. L2:84 weakly excites
  3. L0:2445 weakly excites
  4. L1:1244 weakly suppresses "of" after "board" or "off"
  5. L1:541 weakly excites
  6. L0:1224 weakly suppresses
  7. L0:901 weakly suppresses 'this' after sentence-initial 'In'
  8. L2:1905 weakly excites
  9. L2:2098 weakly suppresses
  10. L2:1036 weakly excites
  11. L0:3017 weakly suppresses
  12. L1:2477 weakly suppresses
  13. L1:337 weakly excites
  14. L0:178 weakly suppresses
  15. L2:872 weakly excites
  16. L2:298 weakly excites
  17. L2:2152 weakly excites
  18. L2:1494 weakly suppresses
  19. L1:1016 weakly suppresses
  20. L0:1301 weakly suppresses
  21. L0:2020 weakly excites
  22. L2:2263 weakly suppresses
  23. L1:767 weakly suppresses
  24. L2:1269 weakly excites

Feeds later layers

  1. L7:42 suppresses
  2. L4:546 suppresses
  3. L8:562 suppresses
  4. L8:376 suppresses
  5. L13:2325 weakly excites
  6. L9:212 weakly excites
  7. L19:1017 weakly suppresses
  8. L12:1851 weakly suppresses hyphen inside dates
  9. L23:2583 weakly excites
  10. L18:1829 weakly excites
  11. L7:1014 weakly suppresses
  12. L7:836 weakly suppresses
  13. L16:1782 weakly suppresses
  14. L6:2170 weakly suppresses
  15. L16:2747 weakly excites
  16. L6:1702 weakly suppresses "not" or "instead" contrasting a correction
  17. L5:51 weakly suppresses
  18. L17:445 weakly suppresses
  19. L10:2046 weakly excites
  20. L5:2657 weakly excites
  21. L16:520 weakly excites
  22. L5:2118 weakly excites
  23. L13:2356 weakly suppresses
  24. L16:1015 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 -6.48 and -6.291 values between -6.29 and -6.101 values between -4.78 and -4.601 values between -4.03 and -3.841 values between -3.47 and -3.281 values between -2.72 and -2.531 values between -1.59 and -1.401 values between -1.40 and -1.213 values between -1.21 and -1.023 values between -1.02 and -0.843 values between -0.84 and -0.659 values between -0.65 and -0.4658 values between -0.46 and -0.27774 values between -0.27 and -0.082442 values between -0.08 and 0.11525 values between 0.11 and 0.2955 values between 0.29 and 0.4819 values between 0.48 and 0.677 values between 0.67 and 0.862 values between 0.86 and 1.05offnegativepositiveoften

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. ichever
  2. 共产党员
  3. 让他们
  4. ATERIAL
  5. 每个月
  6. adesh
  7. 每一个
  8. 례
  9. 洒
  10. ALAR

Suppresses

  1. Others
  2. thur
  3. ых
  4. other
  5. other
  6. _under
  7. loaf
  8. ổi
  9. outros
  10. лез