Neuron Explorer

Qwen3-0.6B

'not' in 'if not most/all' phrases

A neuron in layer 15 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 word 'not' in the phrase 'if not', as in 'most, if not all' or 'if not the most important'. It is the spot where a writer is about to add a stronger or alternative claim. Turning it up pushed the model into looping, repetitive writing.

Written by Claude from this neuron's activations and steering outputs. Confidence: high. 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

    ) version of this image.⏎Description Iris germanica is believed to be the ancestor of many, if not most, of the modern bearded irises that are so popular with gardeners

  2. #2 peak

    three cities: Oklahoma City, El Reno or Enid.⏎At its statehood, Oklahoma was also home to second group generally⏎thought of as German, if not by officers of the United States Bureau⏎of

  3. #3 peak

    is one of the most important, if not the most important, metrics in valuing the bond.⏎Now, let's look

  4. #4 peak

    tanks, large flags, chest thumping marches, discipline-for-itself and martyr walls help inspire nationalism and patriotism? This becomes a particularly good question, if not a surprising one, because, historically speaking, ideas of nationalism and patriotism have never

  5. #5 peak

    eating plan.⏎Beyond promoting every nutrition expert’s mantra of “eat more vegetables and fruits”, the standard raw food diet is true to its name – most, if not all, foods need to be “raw”. This means dieters are prohibited from

  6. #6 peak

    ification which is in faith (1 Tim. 1:4; cf. Titus 3:9). In the modern church, however, many—if not most—believers who are reading Scripture are prone to skip over the genealog

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 think it is called to be called. I think it is called to be called. The term, which is known as to be called. The term, which is known as to be called. The term, which is known as to

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 most common types of tomatoes available. They can be meaty like a beefsteak, but always juicy.⏎Vesuvio Piennolo<|endoftext|>Thank you to DadsWorksheets.com for providing a VARIETY of math practice sheets

  2. #2 peak

    of Frank Churchill’s secret engagement to Jane.⏎“Good God!” cried Emma, “Jane actually on the point of going as governess! What

  3. #3 peak

    in Europe and western Asia, but it was introduced and naturalized to Australia, New Zealand, and North America over time.⏎Want to save this post for later? Click Here to Pin It On Pinterest!⏎Are you interested in learning more about this

  4. #4 peak

    will present new data concerning a stereo photography studio that Dalmau opened in 1854.⏎Download data is not yet available.⏎How to Cite⏎Cuenca, C. (2019). The The “Animated

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. L8:2127 excites
  2. L13:2739 weakly excites
  3. L13:76 weakly suppresses
  4. L14:1436 weakly suppresses 'and' linking a role to a name
  5. L7:2138 weakly suppresses
  6. L12:63 weakly suppresses
  7. L4:1646 weakly suppresses 'fixed' as a finance or cost term
  8. L6:1259 weakly suppresses
  9. L12:1504 weakly suppresses
  10. L1:2950 weakly excites
  11. L10:272 weakly suppresses
  12. L12:1055 weakly suppresses
  13. L5:563 weakly excites
  14. L14:2959 weakly suppresses
  15. L11:1085 weakly excites
  16. L7:823 weakly excites
  17. L9:884 weakly excites
  18. L10:660 weakly suppresses
  19. L6:2120 weakly excites
  20. L8:1235 weakly suppresses
  21. L12:2424 weakly suppresses
  22. L6:2078 weakly suppresses
  23. L5:1322 weakly excites
  24. L5:869 weakly suppresses

Feeds later layers

  1. L19:2863 weakly excites
  2. L21:761 weakly suppresses
  3. L16:815 weakly suppresses 'the' before a noun mid-sentence
  4. L18:855 weakly excites
  5. L21:1445 weakly suppresses
  6. L16:162 weakly suppresses
  7. L20:1560 weakly suppresses
  8. L17:2677 weakly suppresses
  9. L19:2401 weakly excites
  10. L26:2701 weakly excites
  11. L21:2456 weakly excites
  12. L16:271 weakly excites
  13. L22:1773 weakly suppresses
  14. L18:1797 weakly suppresses
  15. L18:1228 weakly suppresses
  16. L20:1051 weakly suppresses
  17. L26:1275 weakly suppresses
  18. L25:501 weakly suppresses
  19. L20:2229 weakly suppresses
  20. L16:801 weakly suppresses
  21. L17:1530 weakly suppresses
  22. L24:2590 weakly suppresses
  23. L18:1679 weakly suppresses
  24. L21:224 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.

6 values between -1.56 and -1.188 values between -1.18 and -0.8154 values between -0.81 and -0.441124 values between -0.44 and -0.062466 values between -0.06 and 0.31145 values between 0.31 and 0.6836 values between 0.68 and 1.0620 values between 1.06 and 1.4314 values between 1.43 and 1.817 values between 1.81 and 2.183 values between 2.18 and 2.554 values between 2.55 and 2.933 values between 2.93 and 3.304 values between 3.30 and 3.673 values between 3.67 and 4.055 values between 4.05 and 4.421 values between 4.42 and 4.792 values between 5.17 and 5.541 values between 5.91 and 6.291 values between 6.29 and 6.661 values between 13.01 and 13.38offnegativepositiveoften

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. BeenCalled
  2. _YES
  3. >true
  4. GetY
  5. oyer
  6. :Boolean
  7. onLoad
  8. /gpl
  9. +=⏎
  10. emple

Suppresses

  1. ials
  2. ино
  3. ports
  4. .nlm
  5. -san
  6. jedoch
  7. imming
  8. imagin
  9. berman
  10. Reviewed