Neuron Explorer

Qwen3-0.6B

'for' after past participles like selected

A neuron in layer 7 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 'for' when it comes right after a verb like 'selected', 'tested' or 'chosen', where it introduces a purpose or reason. Turning it up did not clearly change the writing in a visible way.

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

    in theory, target any protein that you have a ligand for and send it off for destruction. Building these molecules can be a challenge, and the choice of linking groups

  2. #2 peak

    and a Dictionary of controlled vocabulary.⏎The Writing Rules cover aspects of grammar and style.⏎The Dictionary specifies the general words that can be used.⏎These words were chosen for their simplicity and ease of recognition. In general, there is only one word for

  3. #3 peak

    arvesting has threatened its long-term sustainability, and there is a knowledge gap relating to the effect of harvesting on the ecosystem. After conifer species were selected for harvesting, the mixed Chinese fir, pine, and

  4. #4 peak

    summer as a thesis project for my M.A. degree in Anthropology at the University of Alaska. This site was selected for several reasons. First among the reasons for selecting this site was the fact that it

  5. #5 peak

    Everything will be tested for the first time at a Ugandan hospital, the Comprehensive Rehabilitation Services for Uganda (

  6. #6 peak

    and wear resistance. But, we need to always remember that tiles are not indestructible. If mistreated, or subjected to harsh conditions, or wrongly selected for purpose, the glaze on a ceramic tile will scratch or it can wear.⏎

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: 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10,

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

    as a sports arena..⏎Do we have 2 moons?⏎After more than half a century of speculation and controversy, Hungarian astronomers and physicists say they have finally confirmed the existence of

  2. #2 peak

    mathematics, the liberal arts, and the value of a well-rounded education may find themselves echoing Daniel Dennett: “There’s nothing I like less than bad arguments for

  3. #3 peak

    from a quarry and shaped using bronze tools, then smoothed with sand in order to fit tightly together. No mortar was used in the construction. Even after many centuries of wear and tear, the precision is amazing; It would be impossible to slide a

  4. #4 peak

    Korean Peninsula, while the northern part re-emerged as Balhae, a successor-state of Goguryeo. After nearly 1000 years of rule, Silla fragmented into the brief Later Three Kingdoms, Silla,

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. L5:3049 suppresses
  2. L3:381 excites 'for' in research institute names
  3. L6:2379 suppresses
  4. L4:1035 suppresses 'for' after verbs like 'fall' or 'click'
  5. L6:685 excites
  6. L6:368 excites
  7. L6:2718 weakly suppresses
  8. L4:677 weakly suppresses
  9. L6:439 weakly excites
  10. L4:1154 weakly excites
  11. L4:1023 weakly suppresses
  12. L3:159 weakly suppresses
  13. L6:1684 weakly suppresses
  14. L4:1525 weakly suppresses
  15. L5:1090 weakly excites
  16. L4:1013 weakly suppresses
  17. L1:1516 weakly suppresses
  18. L5:2251 weakly excites
  19. L5:2071 weakly suppresses
  20. L4:2748 weakly suppresses
  21. L0:2534 weakly excites
  22. L6:2907 weakly suppresses
  23. L4:1897 weakly suppresses
  24. L6:1880 weakly suppresses "by" introducing a deadline or means

Feeds later layers

  1. L17:3000 excites
  2. L24:480 excites
  3. L21:2338 excites
  4. L16:817 suppresses
  5. L15:2925 excites
  6. L17:1196 excites
  7. L26:298 excites
  8. L20:233 suppresses
  9. L18:622 excites
  10. L8:655 suppresses
  11. L27:303 excites
  12. L27:208 weakly excites
  13. L17:264 weakly suppresses
  14. L20:2948 weakly excites
  15. L9:1269 weakly suppresses
  16. L10:272 weakly suppresses
  17. L9:143 weakly suppresses
  18. L19:1840 weakly excites
  19. L10:459 weakly excites
  20. L18:277 weakly suppresses
  21. L22:1224 weakly suppresses
  22. L8:45 weakly excites
  23. L17:790 weakly excites
  24. L21:1813 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.

3 values between -1.13 and -0.893 values between -0.89 and -0.6536 values between -0.65 and -0.41653 values between -0.41 and -0.172632 values between -0.17 and 0.07454 values between 0.07 and 0.3150 values between 0.31 and 0.5519 values between 0.55 and 0.7915 values between 0.79 and 1.039 values between 1.03 and 1.274 values between 1.27 and 1.514 values between 1.51 and 1.754 values between 1.75 and 1.995 values between 1.99 and 2.235 values between 2.23 and 2.472 values between 2.47 and 2.713 values between 2.71 and 2.952 values between 2.95 and 3.193 values between 3.43 and 3.671 values between 4.63 and 4.871 values between 8.23 and 8.47offnegativepositiveoften

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. ilmington
  2. litigation
  3. áb
  4. ól
  5. 地理位置
  6. cleanliness
  7. 综合治理
  8. kính
  9. ót
  10. inking

Suppresses

  1. verse
  2. vers
  3. deselect
  4. hitch
  5. NSSet
  6. 饰
  7. Harness
  8. utter
  9. .setChecked
  10. หมด