Neuron Explorer

Qwen3-0.6B

'of' or 'the' after 'part of' or 'subset of'

A neuron in layer 14 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 words right after phrases like 'part of', 'subset of' or 'during', where the next word names a larger group or category. Turning it up gave text about other topics, so the output does not clearly reflect this.

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

    against a predetermined set of requirements and validating that the product meets those requirements.⏎Examples of QC include technical reviews, software testing and code inspections.⏎Testing is a subset of QC. It is the process of executing a system in order to detect bugs in

  2. #2 peak

    can have as many base search paths as you like. Here's one that includes⏎/etc as a search path too.<|endoftext|>Salivary gland (Endocrine system)⏎: #1 is , #2 is , #3 is⏎The sal

  3. #3 peak

    floxacin is an antibiotic that is part of the fluoroquinolone family. It effectively kills bacteria, while dexameth

  4. #4 peak

    , Thamserku, Kantega, Mera Peak and Island Peak.⏎Mt. Everest, which is part of the Himalaya range, is located on the border between Nepal and Tibet. Rising to

  5. #5 peak

    However, the Supreme Court justices are chosen by the president, the head of the executive branch. But while the president makes that selection, the Senate, part of the legislative branch, approves the decision.⏎Each branch must depend on the other in order

  6. #6 peak

    <|endoftext|>The Battle of the Kasserine Pass was a battle that took place during the Tunisia Campaign of World War II in February of 1943. It

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 have been, and has been, and is most well known by me. I have been, and is most well known by me, and is most well known by me. 100% of the first person. 1

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

    the 1999 State of the Union Address, and then determine how newspaper articles and television reports describe and analyze the event. Use this...⏎6th - 12th English Language Arts CCSS: Adaptable⏎The State

  2. #2 peak

    in-depth reports from Israel⏎Connect with Israel, right from your home.⏎Lift up the voice of truth and hope⏎Support Jerusalem-based Zionist journalism.⏎Already a member? Login here.<|endoftext|>The fractured rock surfaces were observed under different loading stages

  3. #3 peak

    characters shorter!⏎The current version of bool-extras is 0.3.0 and is available on Hackage.⏎If you have cabal-install (which you really should have, by the way), the

  4. #4 peak

    elements of a successful speech. Not only do they get to become better-informed...⏎7th - 11th Social Studies & History⏎President Obama's State of the Union

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. L12:1356 excites
  2. L13:2420 excites
  3. L8:1885 suppresses
  4. L10:2016 weakly suppresses
  5. L13:3062 weakly suppresses
  6. L9:1737 weakly suppresses
  7. L13:1489 weakly suppresses
  8. L3:647 weakly suppresses
  9. L7:2108 weakly suppresses
  10. L13:2564 weakly suppresses
  11. L9:242 weakly excites
  12. L5:1096 weakly excites
  13. L12:1590 weakly excites
  14. L12:995 weakly suppresses
  15. L1:199 weakly excites
  16. L12:2040 weakly suppresses
  17. L0:338 weakly excites
  18. L6:784 weakly excites
  19. L8:1804 weakly suppresses
  20. L13:2999 weakly suppresses
  21. L2:2516 weakly excites
  22. L5:1477 weakly excites
  23. L12:3006 weakly excites
  24. L12:3017 weakly suppresses

Feeds later layers

  1. L15:2333 weakly excites
  2. L15:978 weakly suppresses
  3. L20:2988 weakly excites
  4. L18:1691 weakly excites
  5. L15:1698 weakly suppresses
  6. L16:1759 weakly suppresses
  7. L18:1858 weakly excites
  8. L19:2443 weakly excites
  9. L15:1095 weakly suppresses
  10. L15:2058 weakly suppresses
  11. L17:2175 weakly excites
  12. L22:2127 weakly suppresses
  13. L16:1121 weakly excites
  14. L16:1584 weakly suppresses
  15. L18:1699 weakly suppresses
  16. L18:1356 weakly excites
  17. L17:909 weakly suppresses
  18. L17:209 weakly excites
  19. L18:1879 weakly suppresses
  20. L15:2221 weakly excites
  21. L18:327 weakly excites
  22. L15:2195 weakly excites
  23. L20:1817 weakly suppresses
  24. L16:2788 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.

4 values between -1.32 and -1.168 values between -1.16 and -1.0014 values between -1.00 and -0.8320 values between -0.83 and -0.6725 values between -0.67 and -0.51103 values between -0.51 and -0.35245 values between -0.35 and -0.19838 values between -0.19 and -0.031969 values between -0.03 and 0.13410 values between 0.13 and 0.29136 values between 0.29 and 0.4549 values between 0.45 and 0.6123 values between 0.61 and 0.7719 values between 0.77 and 0.9312 values between 0.93 and 1.098 values between 1.09 and 1.257 values between 1.25 and 1.425 values between 1.42 and 1.583 values between 1.58 and 1.742 values between 1.74 and 1.901 values between 1.90 and 2.061 values between 2.06 and 2.222 values between 2.22 and 2.382 values between 2.70 and 2.861 values between 3.66 and 3.821 values between 4.95 and 5.11offnegativepositiveoften

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. ialect
  2. @update
  3. extras
  4. ubyte
  5. ungs
  6. NodeType
  7. CIT
  8. toc
  9. egov
  10. avia

Suppresses

  1. ancel
  2. 泣
  3. 干
  4. 精子
  5. åde
  6. edian
  7. `,`
  8. 从
  9. rance
  10. rence