Neuron Explorer

Qwen3-0.6B

the word 'subject' in subject areas

A neuron in layer 24 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 'subject' in phrases like 'ASJC Scopus subject areas' or 'subject expert', meaning a field of study. Turning it up gave text about areas and regions, so the link to the word 'area' shows up in the steering.

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

    pages||12|⏎|Journal||Classical and Quantum Gravity|⏎|State||Published - Jul 7 2003|⏎ASJC Scopus subject areas⏎- Physics and Astronomy (miscellaneous)<|endoftext|>Most computer users are familiar

  2. #2 peak

    controls, as well as in a discontinuity analysis on observations near the borders between anglophone and francophone countries.⏎- Colonial rule⏎ASJC Scopus subject areas⏎- Economics and Econometrics⏎- Organizational Behavior and Human Resource Management

  3. #3 peak

    triple-layer coated and graded-index-layer coated samples was performed and compared with experimentally derived films. Good agreement between model and experiment was observed.⏎ASJC Scopus subject areas⏎- Electronic, Optical and Magnetic Materials⏎- Ceramics and Composites

  4. #4 peak

    NCERT Solutions for Class 10 Maths Chapter 6 Triangles Exercise 6.5 are given here. There are several questions in this exercise. Our subject expert teachers have solved each of these questions in a step by step manner. All

  5. #5 peak

    visual as well as audio learners.⏎BBC Educational Video subject areas⏎For business studies, video brings the complexities of business to life through real

  6. #6 peak

    -popular History section.⏎In our new position on the Lower Ground Floor, next to City Library, in the newly refurbished Central Library, we have retained the same subject sections for

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:iaiaiaiaiaiaiaiaiaiaiaiaiaiaiaiaiaiaiaiaiaiaiaiaiaiaiaiaiaiaiaiaiaiaiaiaiaiaiaiaiaiaiaiaiaiaiaia

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

    are taking it regularly, take it as soon as possible. If it is almost time for your next dose, skip the missed dose and go back to your regular dosing schedule. Do not take 2 doses at once.⏎Store Ampicillin

  2. #2 peak

    of component features. Close to the entrance is a bath complex. The owner had his own private baths. The antechamber leads to a courtyard with columns and walkways. The antechamber is not directly lined up with the axis of the

  3. #3 peak

    been added throughout this edition and rationales for the answers continue to be provided giving the students a better understanding of principles related to drug dosages. This fifth edition addresses the increasing responsibility of nurses in medication and administration;

  4. #4 peak

    - Health Sciences Division⏎Also available at Amazon.com⏎Calculate with Confidence provides a clear consistent format with a step-by-step approach to the calculation and administration of drug dosages. It covers the ratio and proportion, formula, and dimensional analysis methods.

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. L19:314 excites
  2. L16:2743 excites
  3. L20:245 suppresses
  4. L14:1774 excites
  5. L16:507 weakly excites
  6. L22:394 weakly suppresses
  7. L22:1089 weakly excites 'time' in time frame phrases
  8. L3:242 weakly suppresses
  9. L9:153 weakly excites
  10. L17:1271 weakly excites
  11. L18:1321 weakly suppresses
  12. L22:151 weakly suppresses
  13. L19:207 weakly suppresses
  14. L14:2579 weakly suppresses
  15. L17:1241 weakly excites
  16. L7:975 weakly suppresses
  17. L12:1218 weakly suppresses
  18. L18:191 weakly suppresses
  19. L4:737 weakly suppresses
  20. L16:769 weakly excites
  21. L15:1315 weakly excites line break splitting a sentence
  22. L21:508 weakly excites
  23. L21:38 weakly excites
  24. L19:2208 weakly suppresses

Feeds later layers

  1. L26:2216 weakly excites
  2. L27:1910 weakly excites
  3. L27:1721 weakly excites
  4. L27:1074 weakly excites
  5. L25:1921 weakly excites
  6. L25:2950 weakly excites 'in' in 'in progress' phrases
  7. L27:1027 weakly excites
  8. L25:1079 weakly excites
  9. L25:1624 weakly suppresses
  10. L26:1854 weakly suppresses
  11. L27:1757 weakly excites
  12. L25:1470 weakly excites
  13. L25:2443 weakly suppresses
  14. L27:1634 weakly excites
  15. L27:2942 weakly excites
  16. L26:1944 weakly suppresses
  17. L25:2327 weakly suppresses
  18. L25:274 weakly excites
  19. L26:2405 weakly excites
  20. L27:2026 weakly suppresses
  21. L25:1853 weakly excites
  22. L27:1037 weakly excites
  23. L25:3008 weakly excites
  24. L27:1157 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 -28.88 and -27.676 values between -11.98 and -10.771 values between -10.77 and -9.5613 values between -8.36 and -7.155 values between -7.15 and -5.9410 values between -5.94 and -4.7310 values between -4.73 and -3.5324 values between -3.53 and -2.3266 values between -2.32 and -1.112587 values between -1.11 and 0.091022 values between 0.09 and 1.3067 values between 1.30 and 2.5129 values between 2.51 and 3.7228 values between 3.72 and 4.9211 values between 4.92 and 6.1313 values between 6.13 and 7.344 values between 7.34 and 8.542 values between 8.54 and 9.755 values between 9.75 and 10.961 values between 10.96 and 12.161 values between 14.58 and 15.791 values between 16.99 and 18.201 values between 18.20 and 19.41offnegativepositiveoften

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. areas
  2. Areas
  3. Areas
  4. _areas
  5. domains
  6. area
  7. areas
  8. domain
  9. 領域
  10. Area

Suppresses

  1. irs
  2. age
  3. aves
  4. ize
  5. IRS
  6. anness
  7. ite
  8. ia
  9. ensors
  10. ensor