Neuron Explorer

Qwen3-0.6B

'to' in 'X to Y' phrases

A neuron in layer 5 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 responds to the word 'to' in pairs like 'year to year' or 'person to person', where something is compared or passed across. Its opposite side reacts to 'which' after 'from', as in 'from which'. Turning it up did not bring out these phrases in the output.

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

    000. That year, Canadian women had given birth to 479,000 children.⏎- With Canada's population steadily growing from year to year, the number of deaths has understandably increased over time, and it will continue

  2. #2 peak

    however, when James Kirk took command, and ordered a pursuit of the Narada toward Earth. (Star Trek)<|endoftext|>It's just fluctuating wildly from year to year as it always has done but by making some complex "allowances", climate

  3. #3 peak

    either top to bottom or bottom to top). If the trimmer is placed parallel to the nail (cutting from side to side), the nail is crushed and may splinter.⏎The handles of the trim

  4. #4 peak

    comply may result in legal action.⏎WHAT YOU NEED TO KNOW:⏎Diphtheria is a disease caused by a bacterial infection. The infection spreads quickly from person to person through sneezing or coughing. It

  5. #5 peak

    in South Korea. Human-to-human transmission in this outbreak as in all others has been limited, indicating that this MERS-CoV does not easily spread from person to person. The exact source of MERS-CoV is unknown at this time.

  6. #6 peak

    below the 48°N parallel is 474/year with April and May generally being the most active months. The numbers vary quite widely from year to year. For example, in 2011 only three icebergs

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 need to be able to find a way to get a job. I need to be able to get a job. I need to be able to get a job. I need to be able to get a job. I need to be

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

    say the prints date to the middle of the Jurassic Period, a time from which few dinosaur fossils have been recovered.⏎Researchers reported this month on their study of about

  2. #2 peak

    end they don't look at the writer by what he writes, but rather evaluate him by where he lives, or by the location of "the room" from which he writes, as the Peruvian writer Vargas Llosa noted in his

  3. #3 peak

    of Ireland upon the horizon." In Iona accordingly he resolved to make his home. The spot from whence St. Columba made this sorrowful survey is still called by the isles

  4. #4 peak

    from basic waxing and waning of the moon to the arrival of comets and eclipses.⏎The word "space" was specifically mentioned as the place from where the meteorite that made Toph's bracelet and Sokka's sword came.

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. L4:1593 excites
  2. L3:1119 excites
  3. L4:192 excites
  4. L1:2195 excites
  5. L4:1002 weakly suppresses
  6. L2:1015 weakly suppresses
  7. L4:2146 weakly suppresses
  8. L1:2387 weakly excites
  9. L1:2984 weakly suppresses
  10. L2:2450 weakly excites
  11. L3:1079 weakly suppresses
  12. L4:3068 weakly excites
  13. L4:1726 weakly excites
  14. L3:2937 weakly suppresses
  15. L3:807 weakly suppresses comma inside dates and years
  16. L2:1350 weakly excites
  17. L0:2442 weakly suppresses
  18. L3:2372 weakly excites
  19. L3:1736 weakly suppresses
  20. L1:1414 weakly excites
  21. L3:2229 weakly suppresses
  22. L3:2166 weakly suppresses
  23. L0:896 weakly excites
  24. L1:2350 weakly excites 'reasons' or 'permitting' after a qualifier

Feeds later layers

  1. L11:658 strongly suppresses
  2. L9:1382 suppresses
  3. L18:2122 suppresses
  4. L7:1563 suppresses
  5. L16:1283 suppresses
  6. L19:1997 suppresses
  7. L7:1393 suppresses
  8. L10:1054 suppresses 'from' after movement or transfer verbs
  9. L20:1980 weakly suppresses
  10. L7:1533 weakly suppresses
  11. L7:1399 weakly suppresses "to" between a starting and ending value
  12. L6:951 weakly suppresses
  13. L10:1603 weakly suppresses
  14. L13:1329 weakly suppresses
  15. L19:1871 weakly suppresses
  16. L11:2214 weakly suppresses
  17. L17:1651 weakly suppresses
  18. L10:42 weakly suppresses
  19. L13:2955 weakly suppresses
  20. L20:499 weakly suppresses
  21. L7:626 weakly excites
  22. L16:2231 weakly excites
  23. L9:1961 weakly suppresses 'of' in titles and place names
  24. L19:2374 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 -3.15 and -3.051 values between -2.64 and -2.541 values between -1.92 and -1.822 values between -1.61 and -1.511 values between -1.00 and -0.901 values between -0.90 and -0.791 values between -0.79 and -0.693 values between -0.59 and -0.495 values between -0.49 and -0.3813 values between -0.38 and -0.2872 values between -0.28 and -0.18404 values between -0.18 and -0.081228 values between -0.08 and 0.031326 values between 0.03 and 0.13616 values between 0.13 and 0.23173 values between 0.23 and 0.3441 values between 0.34 and 0.4412 values between 0.44 and 0.543 values between 0.54 and 0.643 values between 0.64 and 0.751 values between 0.85 and 0.95offnegativepositiveoften

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. lest
  2. xAD
  3. addPreferredGap
  4. letic
  5. ModelProperty
  6. 场
  7. GROUP
  8. itudes
  9. xes
  10. ♭

Suppresses

  1. oire
  2. 往
  3. Carthy
  4. 年开始
  5. iente
  6. 로부터
  7. ersen
  8. ič
  9. ultan
  10. getInt