Neuron Explorer

Qwen3-0.6B

digits inside years like 2017

A neuron in layer 13 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 digits inside years, especially the zero in years like 2011, 2017 or 2021. It lights up less on years from long ago, like 48 BCE or 1793. When it was turned up, the model's text became rambling and off topic, so the steering output does not show a clear theme.

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

    flu severity, it helps them better understand whether its widespread. An infographic in the article shows just how rampant reports of flu-like symptoms were in late November 2017 as compared to previous years.⏎And while that gives experts plenty of reason

  2. #2 peak

    11 southern hemisphere influenza seasons.⏎During weeks 29 to 30 (17 July 2011 to 30 July 2011), National Influenza Centres (NICs) from

  3. #3 peak

    21 - SARS CoV-2: Its Morphology, Genome, and Implications for Therapy and Vaccines⏎September 29, 2021 - Vaccine Types and Their Effectiveness⏎The recent approval of the commercial

  4. #4 peak

    questions in this exercise. Our subject expert teachers have solved each of these questions in a step by step manner. All these solutions are beneficial for the students of class 10 appearing for the board exams. Students can easily NCERT Solutions for Class

  5. #5 peak

    , 4 PM⏎AuctionsNov 23, 2021, 1 PM⏎World StampsNov 22, 2021, 7 PM<|endoftext|>At SACRE, the Fluent CFD code

  6. #6 peak

    electric power generation. Journal of Hydrology (NZ), 34(1), 1-14.⏎Permanent Research Commons link: http://hdl.handle.net/10289/4701⏎Controlled groundwater abstraction

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: The text is: "The text is: 1004479a6b0a6e6f0a6e6d479f0a6e6d479f0a

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

    Egypt. Simultaneously, she sought to avenge the death of her son Khemu at the hands of the Order of the Ancients, and in 48 BCE, she killed two of their members, Actaeon and Ktes

  2. #2 peak

    as a constitutional monarch and appealed for an end to the Revolution. He retired to Grenoble, and was tried for treason and guillotined (1793). His history of the French Revolution, written during his imprisonment, is considered

  3. #3 peak

    Hidden Ones to resume the fight against the Order, with the former resolving to establish a Roman branch at the expense of her marriage.⏎On 15 March 44 BCE, this culminated in Aya slaying Septimius in personal

  4. #4 peak

    do wolves like it when it is snowing? We are hoping to answer all of our questions through research.⏎The school bus came at 9:30 on a cold, windy, soggy day and picked us up to

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. L11:1629 suppresses
  2. L12:1809 suppresses
  3. L9:349 suppresses the spot before a year
  4. L12:1264 excites gap before a year in dates
  5. L10:1304 suppresses
  6. L11:1698 suppresses
  7. L12:940 suppresses
  8. L11:1246 excites
  9. L12:2401 excites
  10. L8:421 excites
  11. L10:1613 weakly suppresses
  12. L11:2012 weakly suppresses
  13. L12:2338 weakly excites
  14. L11:2739 weakly suppresses
  15. L11:2213 weakly excites
  16. L9:415 weakly excites
  17. L3:2789 weakly suppresses
  18. L12:2163 weakly excites
  19. L10:1470 weakly suppresses
  20. L8:283 weakly suppresses
  21. L0:1464 weakly excites
  22. L0:176 weakly suppresses
  23. L12:2205 weakly suppresses
  24. L12:3030 weakly suppresses

Feeds later layers

  1. L17:1051 suppresses
  2. L14:759 suppresses
  3. L15:2637 suppresses
  4. L15:924 excites
  5. L15:2236 suppresses
  6. L16:3033 weakly suppresses
  7. L18:995 weakly suppresses
  8. L17:52 weakly suppresses
  9. L15:587 weakly excites
  10. L15:674 weakly excites
  11. L17:1241 weakly excites
  12. L14:555 weakly suppresses
  13. L15:870 weakly excites
  14. L18:240 weakly suppresses
  15. L15:958 weakly suppresses
  16. L17:45 weakly excites
  17. L16:2936 weakly excites 'of' after a title or position noun
  18. L14:1558 weakly excites
  19. L16:2203 weakly suppresses
  20. L15:73 weakly suppresses
  21. L16:1020 weakly suppresses
  22. L14:443 weakly suppresses
  23. L15:790 weakly suppresses
  24. L14:785 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 -4.72 and -4.562 values between -4.56 and -4.402 values between -3.78 and -3.631 values between -3.01 and -2.851 values between -2.85 and -2.701 values between -2.54 and -2.391 values between -2.39 and -2.231 values between -2.23 and -2.082 values between -2.08 and -1.921 values between -1.92 and -1.771 values between -1.77 and -1.611 values between -1.61 and -1.452 values between -1.45 and -1.302 values between -1.30 and -1.141 values between -1.14 and -0.995 values between -0.99 and -0.835 values between -0.83 and -0.685 values between -0.68 and -0.5213 values between -0.52 and -0.3743 values between -0.37 and -0.21391 values between -0.21 and -0.061870 values between -0.06 and 0.101150 values between 0.10 and 0.25291 values between 0.25 and 0.4178 values between 0.41 and 0.5627 values between 0.56 and 0.726 values between 0.72 and 0.882 values between 0.88 and 1.032 values between 1.34 and 1.50offnegativepositiveoften

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. ProgressHUD
  2. uder
  3. тин
  4. ymous
  5. obel
  6. .printStackTrace
  7. IBE
  8. ��
  9. usalem
  10. gether

Suppresses

  1. 省
  2. .generated
  3. economy
  4. otos
  5. "_
  6. urchases
  7. "
  8. środ
  9. izers
  10. 犬