Neuron Explorer

Qwen3-0.6B

"the" in "over the past" time spans

A neuron in layer 23 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 word "the" when it comes right after "over" in phrases like "over the past decade". When it was turned up, the model got stuck repeating "last" and "past", which fits the kinds of words it helps predict next.

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

    . “The shellfish results indicate the strength of the growing aquaculture sector.”⏎Supply and demand of farmed oysters have risen exponentially in North Carolina over the past

  2. #2 peak

    colonies are disappearing fast. According to the nonprofit organization Pollinator Partnership, the United States has lost over 50 percent of its managed honeybee colonies over the past 10 years.⏎Pollinator gardens are just one of the many ways

  3. #3 peak

    date backwards, using B.C. dates, to the time of the creation of the world. Provisionally, based on the studies published in Biblical Chronology over the past two years, I have come to the year 3932 B

  4. #4 peak

    by 50% before 2030.⏎This commitment isn’t just good for the planet—it also comes with considerable financial benefits. Numerous studies over the past three decades have proven the value of green building certifications, with a rental premium

  5. #5 peak

    proximity to our brains, many people are worried that they might be associated with an increased risk of brain tumors. There have been around 30 research studies over the past decade that have looked at this issue and most have found that:⏎- People

  6. #6 peak

    breakthrough research is Allan Showler, an entomologist in ARS’s Integrated Farming and Natural Resources Research Unit at Weslaco, Texas.⏎“Over the 116 years since the boll

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: though it is a small country, it is a country that is trying to solve its national crisis that is still very serious and unknown. How can we solve this country's problem? annament annament annament annament

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

    and browse over 110,000 wildlife and science products⏎Multi-currency. Secure worldwide shipping⏎Wildlife, science and conservation since 1985<|endoftext|>A new study recently published showed that for women who are overweight

  2. #2 peak

    solar winds. After decades of working on that theory, a 1999 study reported evidence that the Sun's magnetic field had strengthened greatly since the 1880s, which brought more attention to the question of whether increased solar

  3. #3 peak

    problem is relevant for Professor Finnegan's class at WSU.⏎What textbook is this problem found in?⏎Our data indicates that this problem or a close variation was asked in Chemistry: An Atoms First Approach - Zumdahl Atoms 1

  4. #4 peak

    .11i standard. WPA could be implemented through firmware upgrades on wireless network interface cards designed for WEP that began shipping as far back as 1999. However, since the changes required in the wireless access points (AP

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. L19:1999 strongly excites
  2. L19:2563 strongly excites
  3. L20:1259 strongly excites
  4. L18:620 strongly suppresses
  5. L19:2470 strongly suppresses
  6. L21:1245 strongly excites
  7. L20:2127 excites
  8. L19:2536 excites
  9. L17:1255 suppresses
  10. L21:2207 excites
  11. L14:2453 suppresses
  12. L15:2655 suppresses
  13. L11:403 suppresses
  14. L21:1093 excites
  15. L8:255 suppresses
  16. L18:2538 suppresses
  17. L16:1602 excites
  18. L22:600 excites
  19. L13:2042 excites
  20. L20:990 excites
  21. L15:2575 suppresses
  22. L12:2964 suppresses
  23. L17:2719 excites
  24. L20:2894 excites

Feeds later layers

  1. L27:1562 strongly excites
  2. L25:1275 strongly excites
  3. L26:1094 strongly excites
  4. L24:1770 excites
  5. L27:2033 excites
  6. L24:1194 excites
  7. L24:2353 excites
  8. L27:2197 excites
  9. L24:1794 excites
  10. L25:2931 excites
  11. L25:63 weakly suppresses
  12. L25:2579 weakly suppresses
  13. L26:2398 weakly excites
  14. L27:1086 weakly excites
  15. L27:1335 weakly excites
  16. L26:1914 weakly excites
  17. L27:1046 weakly excites
  18. L26:2514 weakly excites the word 'just' after have or has
  19. L24:1835 weakly excites
  20. L27:786 weakly excites
  21. L26:1525 weakly excites
  22. L27:1387 weakly excites
  23. L26:549 weakly excites
  24. L27:2471 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 -11.27 and -10.471 values between -8.89 and -8.091 values between -8.09 and -7.307 values between -7.30 and -6.515 values between -6.51 and -5.712 values between -5.71 and -4.924 values between -4.92 and -4.137 values between -4.13 and -3.3319 values between -3.33 and -2.5431 values between -2.54 and -1.75111 values between -1.75 and -0.951401 values between -0.95 and -0.162063 values between -0.16 and 0.64130 values between 0.64 and 1.4353 values between 1.43 and 2.2222 values between 2.22 and 3.0211 values between 3.02 and 3.8111 values between 3.81 and 4.605 values between 4.60 and 5.405 values between 5.40 and 6.194 values between 6.19 and 6.984 values between 6.98 and 7.782 values between 7.78 and 8.573 values between 9.36 and 10.163 values between 10.16 and 10.951 values between 13.33 and 14.121 values between 19.68 and 20.47offnegativepositiveoften

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. previous
  2. last
  3. past
  4. previous
  5. 之前的
  6. last
  7. Previous
  8. 之前
  9. Previous
  10. (previous

Suppresses

  1. bcd
  2. 둘
  3. unce
  4. contres
  5. /cmd
  6. eries
  7. 不解
  8. 轸
  9. uales
  10. ->{$