Neuron Explorer

Qwen3-0.6B

the last zero in round numbers

A neuron in layer 0 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 a zero inside a number, such as the zeros in years like 2000 or amounts like $5,000. When it was turned up, the written text did not show a clear change in topic or style.

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

    ten years.<|endoftext|>BBB - A+ Rating - the best there is⏎An excerpt from www.HouseOfNames.com archives copyright © 2000 - 2014⏎Where did the Scottish Olyphant family come

  2. #2 peak

    anesthesia in<|endoftext|>BBB - A+ Rating⏎- the best there is⏎An excerpt from www.HouseOfNames.com archives copyright © 2000 - 2014⏎Origins Available: English, Irish⏎Where

  3. #3 peak

    but they say false positives can be reduced by focusing on high-risk patients and using regular instead of digital mammogram's.⏎All content © Copyright 2000 - 2013 WorldNow and WXOW. All Rights Reserved.⏎

  4. #4 peak

    Sumerian civilization is the oldest civilization known to mankind. The term Sumer is today used to designate southern Mesopotamia. In 3000 BC, a flourishing urban civilization existed. The Sumerian civilization was

  5. #5 peak

    including any advertising you do for your business.⏎You may be proud to tell anyone who asks that you’ve processed $50,000 worth of orders, but in reality that number is meaningless. How much did it

  6. #6 peak

    the cosmos! Each day a different image or photograph of our fascinating universe is featured, along with a brief explanation written by a professional astronomer.⏎2000 April 21⏎Explanation: Star formation occurs at a faster pace in M

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 thinking about lately. I have been thinking about lately. I have been thinking about lately. I have been thinking about lately. I have been thinking about lately. I have been thinking about lately. I have been thinking about lately

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

    play a role in determining the threshold for a fear response. There is considerable variation in the canine population regarding the types of responses that are generated by fear-provoking stimuli. Some dogs require a very strong stimulus to elicit fear, while others

  2. #2 peak

    readers, evoking a range of emotions and inspiring self-reflection. [Insert Poem Title] is no exception. Through her vivid descriptions and thought-provoking themes, Oliver encourages readers to pause, appreciate the beauty of nature, and

  3. #3 peak

    constructed to invoke such libraries. In other words, a tag library defines a set of Java classes for the custom JSP actions that the tag library supports. Invoking the tag library imports custom processing and actions into the Java environment so a Java runtime

  4. #4 peak

    that the poems must follow a format suitable for Twitter–280 characters or fewer. These bite-sized verses often prove interesting, complex, and thought-provoking. Listen to this story to hear poet Jessica Care Moore select and read some of

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

None: this is the first layer.

Feeds later layers

  1. L2:1594 excites the number '100' in numerals
  2. L4:308 excites
  3. L19:574 weakly excites
  4. L4:2938 weakly excites
  5. L16:2056 weakly suppresses
  6. L19:265 weakly excites
  7. L4:667 weakly suppresses
  8. L20:2921 weakly suppresses
  9. L2:400 weakly excites
  10. L26:1374 weakly suppresses
  11. L6:740 weakly excites
  12. L12:1265 weakly suppresses
  13. L5:1957 weakly suppresses
  14. L25:1801 weakly excites
  15. L3:2126 weakly suppresses
  16. L1:6 weakly suppresses
  17. L1:1160 weakly suppresses
  18. L25:1234 weakly excites
  19. L9:112 weakly excites
  20. L15:1364 weakly excites
  21. L7:201 weakly suppresses
  22. L6:2398 weakly suppresses
  23. L24:2141 weakly excites
  24. L6:2745 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.

6 values between -0.42 and -0.3211 values between -0.32 and -0.2260 values between -0.22 and -0.12747 values between -0.12 and -0.022707 values between -0.02 and 0.08344 values between 0.08 and 0.1815 values between 0.18 and 0.282 values between 0.38 and 0.483 values between 0.48 and 0.581 values between 0.77 and 0.871 values between 1.87 and 1.972 values between 2.97 and 3.071 values between 3.07 and 3.172 values between 3.27 and 3.373 values between 3.37 and 3.473 values between 3.47 and 3.57offnegativepositiveoften

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. urable
  2. burgh
  3. licher
  4. 牍
  5. rowser
  6. ลบ
  7. 专业人士
  8. esses
  9. istration
  10. �件

Suppresses

  1. nhất
  2. 再
  3. los
  4. NULL
  5. Op
  6. ional
  7. 科学院
  8. 荷
  9. aits
  10. ationship