Neuron Explorer

Qwen3-0.6B

"this" pointing back to a thing

A neuron in layer 14 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 "this" (or "these") when it points back to something just mentioned, like "this converter" or "these proteins". When it was turned up, the output did not show that pattern clearly and became repetitive and unrelated.

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

    sub-processor called a GPU (graphics processing unit). All the calculations that are needed to turn the computer code into the picture on the screen are given over to this second sub processor and it is dedicated to this task. This hand-off of this

  2. #2 peak

    to cook something slowly and gently in water that is just below the boiling point. Usually eggs are what's being cooked when "coddle" is used in this sense today, but when "coddle" was first being

  3. #3 peak

    by...⏎Yersiniae cause severe intestinal infections. Studies using Yersinia pseudotuberculosis as a model organism aim to elucidate the infection mechanisms of these...⏎Researchers from the

  4. #4 peak

    3D printers. They often use 8mm lead screws as demonstrated in the video below. The concepts explained above are all shown in the first few seconds of this video:<|endoftext|>Tourette's Syndrome (cont.)⏎IN THIS ARTICLE⏎Treatment

  5. #5 peak

    (Anderson) saw potential in his work, and began mass-marketing them. He created several versions over the years, the last and most popular of which was this one created in 1941.⏎Christ at Heart's Door. The

  6. #6 peak

    have to fill in the value of Arabic numeral and it will show the same value in Roman numeral. You can convert any desired Arabic numeral to Roman format by using this converter. You can start from number 1 and go up to any possible number

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 to be able to to get the the to get to the to get to to get to. I have to to get to to. I have to to. I have to. I have to. I have to. 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

    1991-1995⏎The materials in this archive are copyright © The Manitoba Library Association. Reproduction for personal use is permitted only if this copyright notice is maintained. Any other reproduction is prohibited without permission Copyright information for reviewers

  2. #2 peak

    three arcs above and to the right of the cluster core, and the fourth is down and to the left, partially obscured by a foreground star. Each arc is that

  3. #3 peak

    and animals lost to the eruptions. The government will buy endangered cows on the volcano, the president said.⏎Many of those who live on its slopes risked their lives by staying or returning to feed their cows during lulls in the volcanic activity

  4. #4 peak

    . How quickly you respond to your computer's symptoms can make all the difference in how critically it is affected. Initially, don't worry about identifying the culprit. That will come later. For now, focus on isolating the problem before it does

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. L0:1035 excites
  2. L5:125 excites "of" after all three, both, ten
  3. L10:167 suppresses
  4. L6:1801 weakly suppresses
  5. L10:965 weakly suppresses the word 'that' introducing a clause
  6. L10:1404 weakly excites
  7. L12:521 weakly excites
  8. L2:602 weakly excites
  9. L12:1065 weakly excites
  10. L9:493 weakly suppresses
  11. L6:1218 weakly suppresses
  12. L4:52 weakly excites
  13. L9:249 weakly excites
  14. L6:61 weakly suppresses
  15. L8:217 weakly suppresses
  16. L8:636 weakly suppresses
  17. L4:968 weakly excites
  18. L2:26 weakly suppresses
  19. L4:731 weakly excites
  20. L1:1318 weakly excites
  21. L4:468 weakly excites "that" after "than" or "like"
  22. L3:229 weakly excites
  23. L6:1135 weakly suppresses
  24. L7:224 weakly suppresses

Feeds later layers

  1. L16:284 excites
  2. L17:985 excites
  3. L15:795 suppresses
  4. L19:832 excites
  5. L20:271 suppresses
  6. L18:1228 excites
  7. L21:511 excites
  8. L21:592 excites
  9. L18:602 excites
  10. L18:1278 excites 'a' or 'most' before a quantity
  11. L16:2520 excites
  12. L15:736 excites
  13. L15:1746 excites
  14. L23:203 excites
  15. L18:77 excites
  16. L17:1027 excites
  17. L22:194 suppresses
  18. L17:1592 suppresses
  19. L16:1711 excites
  20. L20:337 suppresses
  21. L18:141 excites
  22. L22:901 excites
  23. L19:1707 excites
  24. L21:124 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 -2.97 and -2.731 values between -2.49 and -2.252 values between -1.78 and -1.541 values between -1.54 and -1.304 values between -1.30 and -1.075 values between -1.07 and -0.8321 values between -0.83 and -0.59131 values between -0.59 and -0.35893 values between -0.35 and -0.122347 values between -0.12 and 0.12355 values between 0.12 and 0.3672 values between 0.36 and 0.6024 values between 0.60 and 0.8313 values between 0.83 and 1.078 values between 1.07 and 1.313 values between 1.31 and 1.557 values between 1.55 and 1.784 values between 1.78 and 2.024 values between 2.02 and 2.264 values between 2.26 and 2.501 values between 2.50 and 2.731 values between 2.73 and 2.972 values between 2.97 and 3.211 values between 3.92 and 4.161 values between 4.16 and 4.401 values between 4.87 and 5.111 values between 6.29 and 6.53offnegativepositiveoften

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. shortened
  2. ened
  3. ening
  4. two
  5. 复杂的
  6. huge
  7. strange
  8. extensive
  9. particular
  10. four

Suppresses

  1. рам
  2. iota
  3. 看电影
  4. arium
  5. -go
  6. /movie
  7. 纳
  8. èmes
  9. lessness
  10. cci