Neuron Explorer

Qwen3-0.6B

day number after a month name

A neuron in layer 12 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 day number that comes right after a month name, like the 15 in "June 15" or the 28 in "May 28th". When it was turned up, the written text did not show a clear change in topic or style.

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

    , biotech⏎companies activate a patients’ own T cells so that they attack cancer cells.⏎Honoring International Researchers on World Cancer Day⏎Every year on February 4, World Cancer Day unites the world’s population in the fight against cancer

  2. #2 peak

    and decision-making scenarios to help define the process that people go through when formulating solutions to problems.⏎Herbert Simon was born in Milwaukee, Wisconsin, on June 15th, 1916. He went to the Milwaukee public

  3. #3 peak

    . The staff housing located not far from the main buildings are not secluded<|endoftext|>Do you know about Menstrual Hygiene Day, which is observed annually on May 28th?⏎This Saturday is Menstrual Hygiene Day, which will

  4. #4 peak

    appropriate menstrual hygiene management around the world. Menstrual Hygiene Day (MHD, MH Day in short) is a yearly basis awareness day held on May 28 to emphasize the need for excellent menstrual hygiene management (

  5. #5 peak

    everyone in our country because it is not a normal person history it is legend’s biography that should be read by all.ere we are giving short speech on October 2nd. Gandhiji full name is Mohandas Karam Chand Gandhi he

  6. #6 peak

    was born in Montpelier, Vermont, on December 26, 1837. A hero of the Civil War and

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" (myself) in the past, the past. [inserted in the past, but not sure where I was at the moment." (myself) [in the past" ... [end of this passage] 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

    Indus Valley Civilisation of Harappa and Mohenjo-daro from the Bronze Age. Nowadays, the Thar

  2. #2 peak

    statue from Nimrud’s Temple of Ishtar, decorated with cuneiform engravings detailing the life of King Ashurnasirpal II, who reigned from 883 to 8

  3. #3 peak

    - Who produced the first really useful TV system?⏎Answers of Popular Inventions Quiz Questions⏎- William Caxton⏎- The Thermometer and the Telescope⏎- Paper, Printing, Gunpow

  4. #4 peak

    ⏎The major mountains are the Mt.Everest, Mt.Lhotse, Cho Oyu, Nuptse, Pumori, Ama Dablam

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. L11:974 excites
  2. L11:12 weakly suppresses
  3. L9:181 weakly suppresses
  4. L11:1123 weakly excites
  5. L10:2991 weakly excites
  6. L5:509 weakly excites the word 'even' after a clause
  7. L9:2680 weakly excites
  8. L11:2427 weakly excites
  9. L2:497 weakly excites
  10. L9:643 weakly suppresses
  11. L9:171 weakly suppresses
  12. L8:2647 weakly excites
  13. L8:824 weakly excites
  14. L8:1554 weakly excites
  15. L11:54 weakly suppresses
  16. L5:21 weakly excites
  17. L4:2263 weakly excites
  18. L10:1847 weakly excites
  19. L0:1829 weakly suppresses
  20. L7:2362 weakly suppresses
  21. L11:383 weakly excites
  22. L7:1587 weakly excites
  23. L10:2218 weakly suppresses
  24. L10:2117 weakly suppresses

Feeds later layers

  1. L14:2025 excites
  2. L13:1215 excites
  3. L14:2198 excites
  4. L13:291 suppresses
  5. L14:2992 excites
  6. L13:1380 weakly excites
  7. L13:1166 weakly suppresses
  8. L13:1991 weakly excites
  9. L17:2523 weakly excites
  10. L16:8 weakly suppresses
  11. L16:498 weakly excites
  12. L16:83 weakly suppresses
  13. L13:888 weakly suppresses
  14. L14:1038 weakly suppresses
  15. L23:348 weakly excites
  16. L14:1983 weakly suppresses
  17. L16:108 weakly suppresses
  18. L15:45 weakly suppresses
  19. L16:3062 weakly excites
  20. L16:137 weakly suppresses
  21. L13:1966 weakly excites
  22. L16:142 weakly suppresses
  23. L16:44 weakly suppresses
  24. L14:1131 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.

2 values between -3.16 and -3.003 values between -2.70 and -2.541 values between -2.54 and -2.392 values between -2.39 and -2.241 values between -2.08 and -1.932 values between -1.93 and -1.785 values between -1.78 and -1.623 values between -1.62 and -1.479 values between -1.47 and -1.3213 values between -1.32 and -1.1616 values between -1.16 and -1.0112 values between -1.01 and -0.8615 values between -0.86 and -0.7039 values between -0.70 and -0.5553 values between -0.55 and -0.40130 values between -0.40 and -0.24571 values between -0.24 and -0.092032 values between -0.09 and 0.06635 values between 0.06 and 0.22154 values between 0.22 and 0.3785 values between 0.37 and 0.5249 values between 0.52 and 0.6824 values between 0.68 and 0.8316 values between 0.83 and 0.989 values between 0.98 and 1.146 values between 1.14 and 1.296 values between 1.29 and 1.445 values between 1.44 and 1.603 values between 1.60 and 1.751 values between 1.75 and 1.901 values between 1.90 and 2.061 values between 2.06 and 2.212 values between 2.36 and 2.521 values between 2.52 and 2.671 values between 2.82 and 2.98offnegativepositiveoften

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. ures
  2. erah
  3. URES
  4. ImageSharp
  5. egra
  6. 标准
  7. Di
  8. iram
  9. 缁
  10. IVO

Suppresses

  1. doubt
  2. stakes
  3. itet
  4. ไหว
  5. ัย
  6. 糙
  7. át
  8. thead
  9. Textbox
  10. taking