Neuron Explorer

Qwen3-0.6B

word pieces inside names and terms

A neuron in layer 23 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 a middle chunk of a longer name or word that the model has split into pieces, like the 'l' in a surname or the 'elt' in a place name. When it was turned down, the model kept coming back to the words 'objective' and 'goal'. Turned up, the writing just wandered off topic.

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

    large open air theater built in a semicircular pattern, with ascending rows of seats built into a hillside<|endoftext|>Posts tagged education⏎Education Week writer Larry Ferlazzo reached out to me and three other Education & Mental Health professionals to help respond

  2. #2 peak

    from taking her place in the sky. What might she have accomplished had she lived to a ripe old age?<|endoftext|>Sixteen years ago, Patricia L. Gerbarg, a Kingston, N.Y., psychiatrist, contracted Lyme disease. But as

  3. #3 peak

    1416 and initially worked as a Greek scribe. His study under the famous schoolmaster Vittorino da Feltre gave him a mastery of Latin rare among Greek emigres, and by

  4. #4 peak

    Arctic Circle.⏎The ICEHOTEL in Swedish Lapland is celebrating its 20th year.<|endoftext|>Source: Video and Images Created by Amanda Soderlind⏎Welcome to this lesson today on autosomal recessive traits and disorders.

  5. #5 peak

    “In the West, we’re not used to the idea that one drug — or herb — can have a broad range of physical and mental health benefits,” Gerbarg explains. But when you consider that stress has been linked to a host of

  6. #6 peak

    icit that human-like behavior?”⏎Pierson is joined on the paper by Daniela Rus, the Viterbi Professor of Electrical Engineering and Computer Science; Sertac Karaman, associate professor of aeronautics and astronautics;

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 an objective and an objective of the goal. I have an objective of the objective. I have an objective of the objective. I have an objective of the objective. The object of the objective is the object of the objective. I

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

    they should be conducting⏎- How to apply the results to create effective, high performing designs⏎The session will be in English language.<|endoftext|>Statement of Purpose⏎THE CIVIL WAR CENTER⏎The Civil War was arguably the most important event in

  2. #2 peak

    Jepson Interchange (more information)⏎©Copyright 1993 by the Regents of the University of California⏎AND IS MAINTAINED FOR ARCHIVAL PURPOSES ONLY⏎Perennial to trees, from membranous bulb, fib

  3. #3 peak

    pollution.⏎WiseGeek has one of the clearest definitions of Cap and Trade: A cap and trade system is a method for managing pollution, with the end goal of reducing the overall pollution in a nation, region, or industry. Many

  4. #4 peak

    Seven years ago Albert Moore died."⏎Scanned image and text by George P. Landow⏎[This image may be used without prior permission for any scholarly or educational purpose.]<|endoftext|>Tough Task for High School Students: Not all of us are

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. L22:1483 strongly excites
  2. L21:2060 strongly excites
  3. L19:1767 strongly suppresses
  4. L18:1028 excites
  5. L22:2793 excites
  6. L20:2026 excites
  7. L21:2211 suppresses
  8. L21:2168 excites
  9. L22:1725 suppresses
  10. L19:2676 excites
  11. L12:1633 suppresses
  12. L5:2027 suppresses
  13. L20:1602 excites
  14. L18:2659 suppresses
  15. L22:2856 excites
  16. L20:2589 suppresses
  17. L13:1753 suppresses
  18. L6:1737 weakly suppresses
  19. L12:1477 weakly suppresses
  20. L11:2117 weakly excites
  21. L16:2095 weakly excites
  22. L22:2630 weakly excites
  23. L18:2747 weakly suppresses
  24. L12:2450 weakly suppresses

Feeds later layers

  1. L27:2772 strongly suppresses
  2. L25:2948 strongly suppresses
  3. L27:1039 strongly suppresses
  4. L27:1845 strongly suppresses
  5. L24:1675 suppresses
  6. L27:2053 suppresses
  7. L26:2714 weakly suppresses
  8. L24:2784 weakly suppresses
  9. L25:3019 weakly suppresses
  10. L25:1403 weakly suppresses
  11. L25:2611 weakly suppresses
  12. L26:2629 weakly suppresses
  13. L27:3050 weakly suppresses
  14. L25:953 weakly suppresses
  15. L25:1602 weakly suppresses
  16. L24:1282 weakly suppresses
  17. L26:1969 weakly excites
  18. L25:1473 weakly suppresses
  19. L25:2563 weakly excites
  20. L26:1731 weakly suppresses
  21. L25:1601 weakly excites
  22. L24:2077 weakly excites
  23. L24:1963 weakly suppresses
  24. L24:2807 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 -12.40 and -11.931 values between -11.93 and -11.471 values between -10.54 and -10.081 values between -9.15 and -8.692 values between -7.29 and -6.831 values between -6.83 and -6.371 values between -6.37 and -5.904 values between -5.90 and -5.442 values between -5.44 and -4.974 values between -4.97 and -4.512 values between -4.51 and -4.055 values between -4.05 and -3.587 values between -3.58 and -3.126 values between -3.12 and -2.659 values between -2.65 and -2.1916 values between -2.19 and -1.7318 values between -1.73 and -1.2642 values between -1.26 and -0.80179 values between -0.80 and -0.331573 values between -0.33 and 0.131521 values between 0.13 and 0.60419 values between 0.60 and 1.0668 values between 1.06 and 1.5212 values between 1.52 and 1.993 values between 1.99 and 2.452 values between 2.45 and 2.922 values between 2.92 and 3.381 values between 3.38 and 3.841 values between 4.31 and 4.772 values between 5.24 and 5.702 values between 5.70 and 6.16offnegativepositiveoften

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. шив
  2. @$
  3. pond
  4. iless
  5. tục
  6. kar
  7. 湾区
  8. ру
  9. CodeAt
  10. ollapsed

Suppresses

  1. 目的
  2. objective
  3. goal
  4. Objective
  5. purpose
  6. Objective
  7. objectives
  8. objective
  9. aim
  10. 目标