Neuron Explorer

Qwen3-0.6B

opening paren before an acronym

A neuron in layer 13 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 lights up on the opening bracket after a full name, just before its short form, like the "(" in "Graduate Record Examination (GRE)". When it was turned up, the written text did not clearly change in any way that matches this pattern.

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 search of their lost ones.<|endoftext|>The Verbal Reasoning section of the Graduate Record Examination (GRE) challenges a student’s reading comprehension, vocabulary, and sentence completion skills.

  2. #2 peak

    numbers. The app will output the dataset's stem-and-leaf plot once Submit is clicked.<|endoftext|>Americans with Disabilities Act (ADA)⏎The Americans with Disabilities Act (ADA) provides civil rights protections to individuals with disabilities similar to those provided to individuals

  3. #3 peak

    an intact fibula<|endoftext|>Occupational Safety and Health Administration⏎OCCUPATIONAL SAFETY AND HEALTH ADMINISTRATION⏎The mission of the Occupational Safety and Health Administration (OSHA) is to assure that every U.S. worker goes home whole and

  4. #4 peak

    Social Cognition in Non-human Primates Using a Network of Computerized Automated Learning Device (ALDM) Test Systems⏎Institutions: Aix-Marseille University.⏎Fag

  5. #5 peak

    debt to women in dictatorships.”⏎Now, the CIDH has brought the ESMA exhibit to the Organization of American States (OAS) Headquarters in Washington DC in an attempt to start a conversation about crimes

  6. #6 peak

    - Reduce cholesterol and triglyceride levels⏎- Lower blood pressure⏎- Improve memory and cognition⏎What about cinnamon and HIV⏎Acquired immunodeficiency syndrome (AIDS) is caused by HIV and HIV-1 protease plays an essential

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: the same. the same. the same. the same. the same. the same. the same. the same. the same. the same. the same. the same. I have been thinking about, the same. the same.

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

    ://ecf.hq.eso.org/newsletter/stecf-nl-22/freud

  2. #2 peak

    and a copy of the report go to the NOAA news website www.noaanews.noaa.gov⏎– Erich Luening⏎Print this page<|endoftext|>Today is Sunday,

  3. #3 peak

    and numbers. It worked by filling slots in frames maintaining an agenda of resource-limited prioritised tasks. AM's successor was Eurisko. http://homepages.enterprise.net/hibou/aicourse/lenat.txt.⏎Last updated:

  4. #4 peak

    Animal Adaptations". Please register for the second session through Eventbrite as well, if you are interested in attending both: https://scisatadapt2.eventbrite.ca⏎Please contact us if you have any questions & we hope to

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. L9:792 strongly excites
  2. L2:504 strongly suppresses
  3. L0:670 suppresses
  4. L4:300 excites
  5. L9:2661 excites
  6. L8:143 suppresses
  7. L10:321 suppresses
  8. L12:1058 suppresses
  9. L5:1304 excites
  10. L2:1473 excites
  11. L7:366 excites
  12. L6:972 excites "well" after "how"
  13. L10:1979 excites
  14. L8:386 suppresses
  15. L0:860 excites chinese 的 particle before a noun
  16. L5:1851 weakly excites
  17. L0:1523 weakly excites
  18. L3:194 weakly suppresses
  19. L0:339 weakly suppresses
  20. L3:2168 weakly excites
  21. L7:1560 weakly suppresses
  22. L9:1009 weakly suppresses
  23. L1:1215 weakly excites
  24. L9:189 weakly excites

Feeds later layers

  1. L15:526 excites
  2. L17:1999 excites
  3. L15:779 suppresses
  4. L18:1129 suppresses
  5. L14:935 excites
  6. L16:2623 excites
  7. L15:309 suppresses
  8. L15:2616 suppresses
  9. L18:854 excites
  10. L20:2278 excites
  11. L22:1425 excites
  12. L15:2642 excites
  13. L19:2911 excites
  14. L25:2331 excites
  15. L24:140 excites
  16. L16:1433 weakly suppresses
  17. L14:2485 weakly excites
  18. L18:1575 weakly excites
  19. L21:2163 weakly excites
  20. L16:1594 weakly suppresses
  21. L16:1217 weakly suppresses
  22. L17:1996 weakly suppresses
  23. L17:2202 weakly excites
  24. L19:274 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 -1.40 and -1.351 values between -1.18 and -1.131 values between -1.02 and -0.961 values between -0.80 and -0.741 values between -0.74 and -0.691 values between -0.63 and -0.574 values between -0.52 and -0.461 values between -0.46 and -0.416 values between -0.41 and -0.3513 values between -0.35 and -0.3028 values between -0.30 and -0.2478 values between -0.24 and -0.19161 values between -0.19 and -0.13298 values between -0.13 and -0.08518 values between -0.08 and -0.02867 values between -0.02 and 0.04644 values between 0.04 and 0.09478 values between 0.09 and 0.15334 values between 0.15 and 0.20210 values between 0.20 and 0.26137 values between 0.26 and 0.3165 values between 0.31 and 0.3731 values between 0.37 and 0.429 values between 0.42 and 0.489 values between 0.48 and 0.536 values between 0.53 and 0.592 values between 0.64 and 0.701 values between 0.70 and 0.762 values between 0.76 and 0.81offnegativepositiveoften

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. 的过程中
  2. 的真实
  3. 地区的
  4. 产品的
  5. 的时间里
  6. 中的
  7. 里的
  8. 的真实性
  9. .db
  10. 的地

Suppresses

  1. changer
  2. raction
  3. ermal
  4. urret
  5. ategorical
  6. asio
  7. arent
  8. uchs
  9. udent
  10. واء