Neuron Explorer

Qwen3-0.6B

course and business document nouns

A neuron in layer 2 of Qwen3-0.6B. Its strongest snippets are mixed: they peak on different tokens. It is featured, so it has steering outputs.

What this neuron does

This neuron fires on words that show up in course outlines, business guides and reports, such as 'case study', 'action items', 'checklist', 'deliverables' and 'stakeholders'. Turning it up made the model drift toward business and team-process writing, though only loosely.

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

    difficult challenges, and you will discover new ways of looking at your world and yourself--revealing previously hidden opportunities.⏎The book offers real-life stories, explicit action items, and concrete methods that allow you to attain a deeper understanding of any issue,

  2. #2 peak

    Photographer of the Majesty, Light and Space of the Medieval Cathedrals of England and France" in 1973 (Aperture).⏎|[Checklist]||Click<|endoftext|>The unmanned Jules Verne ATV cargo ship breaks up in

  3. #3 peak

    the difference.⏎Quality assurance is process oriented. It is all about preventing defects by ensuring the processes used to manage and create deliverables works. Not only does it work, but is consistently followed by the team.

  4. #4 peak

    Canada, but rather through the initiative of the industry and the relevant ENGOs. This case study investigates the resolution of the conflict. Questions remain as to precisely how this was achieved

  5. #5 peak

    of the agreement. It raises the question of the applicability of the B.C. experience to other fisheries in Canada and the rest of the world.⏎This case study concerns an important component of British Columbia’s commercial capture (wild) fishery,

  6. #6 peak

    viability of their functions, plans and projects.⏎• Manage and control their budget effectively on a day-to-day basis.⏎Introduction, Course outline, Expectations and Takeaways⏎• Understanding the importance of financial statements⏎Analysing & Tools for Pre

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 sum1, A is, the function the total, in (A function) the effect that the effect of the function (A function) the effect that the function (A function) the effect that the function (A function) the

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

    Harvard T.H. Chan School of Public Health, Sept. 2012, https://www.hsph.harvard.edu/nutritionsource/vitamins/.⏎- Betta Health

  2. #2 peak

    . In essence, work management defines the ways organizations manage and improve upon their internal processes to maximize success.⏎Adopting a work management framework means examining the entire range of business operations, identifying the strengths and weaknesses of current processes, and improving them

  3. #3 peak

    .⏎Lake Oroville, Then and Now July 2011 (Paul Hames/CA DWR) August 2014 (Justin Sullivan/Getty) Despite some stormy December days and predictions in Northern California for a wet weekend

  4. #4 peak

    program that strengthens the pigs’ immunity and disease defenses. Producers should identify key interactions where there is a risk of exposure and ensure safety standards are in place. Proper cleaning with disinfectants, heating and drying helps to destroy the virus. Some

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. L1:719 weakly suppresses
  2. L1:101 weakly excites
  3. L0:2920 weakly suppresses
  4. L1:2180 weakly suppresses
  5. L1:1799 weakly excites
  6. L0:392 weakly suppresses
  7. L0:2071 weakly excites
  8. L0:1800 weakly suppresses
  9. L0:2131 weakly suppresses
  10. L1:166 weakly suppresses
  11. L1:2229 weakly suppresses
  12. L1:888 weakly suppresses
  13. L0:2196 weakly excites
  14. L0:555 weakly excites
  15. L0:419 weakly excites
  16. L1:1973 weakly suppresses
  17. L1:2160 weakly suppresses
  18. L0:2485 weakly excites
  19. L1:1527 weakly suppresses
  20. L1:543 weakly suppresses
  21. L1:1608 weakly excites
  22. L1:1663 weakly excites
  23. L0:3037 weakly excites
  24. L1:2146 weakly excites

Feeds later layers

  1. L3:727 excites
  2. L22:1485 excites
  3. L4:2398 weakly excites
  4. L5:173 weakly suppresses
  5. L20:1140 weakly excites
  6. L24:1063 weakly excites
  7. L19:1961 weakly excites
  8. L3:314 weakly suppresses
  9. L18:2657 weakly excites
  10. L21:2721 weakly excites
  11. L6:2007 weakly excites
  12. L3:862 weakly excites
  13. L9:2540 weakly excites
  14. L4:2932 weakly suppresses
  15. L10:763 weakly excites
  16. L24:1474 weakly excites
  17. L21:1363 weakly excites
  18. L9:2938 weakly suppresses
  19. L14:682 weakly suppresses
  20. L8:1348 weakly suppresses
  21. L4:2527 weakly excites
  22. L16:1295 weakly suppresses
  23. L21:2771 weakly excites
  24. L21:1385 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 -1.05 and -0.991 values between -0.99 and -0.931 values between -0.93 and -0.872 values between -0.75 and -0.692 values between -0.69 and -0.634 values between -0.63 and -0.576 values between -0.57 and -0.515 values between -0.51 and -0.453 values between -0.45 and -0.395 values between -0.39 and -0.3319 values between -0.33 and -0.2740 values between -0.27 and -0.21119 values between -0.21 and -0.15248 values between -0.15 and -0.09682 values between -0.09 and -0.031638 values between -0.03 and 0.03663 values between 0.03 and 0.09230 values between 0.09 and 0.15114 values between 0.15 and 0.2152 values between 0.21 and 0.2723 values between 0.27 and 0.3314 values between 0.33 and 0.396 values between 0.39 and 0.459 values between 0.45 and 0.518 values between 0.51 and 0.574 values between 0.57 and 0.631 values between 0.63 and 0.693 values between 0.69 and 0.751 values between 0.93 and 0.991 values between 0.99 and 1.051 values between 1.11 and 1.172 values between 1.29 and 1.35offnegativepositiveoften

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. playbook
  2. ��作
  3. checklist
  4. 妯
  5. tml
  6. aries
  7. (exec
  8. committees
  9. ary
  10. itionally

Suppresses

  1. 势
  2. igar
  3. 剐
  4. ashion
  5. .misc
  6. 怎么样
  7. สะสม
  8. 眇
  9. ede
  10. rades