Neuron Explorer

Qwen3-0.6B

"way" in "find their way"

A neuron in layer 2 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 "way" in phrases like "found his way home" or "find its way into". Turning it up made the model produce unrelated text, so the steering did not reveal the pattern.

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

    product of today’s waste. Most of these are then burned, which means they release toxic chemicals in the air. The one’s that are recycled, find their way back to retailers, who re-seal them, and claim the water is “

  2. #2 peak

    coolant moderator combination had to be one that did not absorb neutrons to any great extent. To make natural uranium “burn” every available neutron has to find its way into fuel material. Using a neutron absorbing coolant/moderator – like the ordinary water

  3. #3 peak

    made to put in a crop for his captor, who conveniently neglected to tell him when the war had ended. Fortunately, Walker escaped again and eventually found his way back home, traveling from Columbus, TX, to Galveston, to New

  4. #4 peak

    . They exist in parts of Arkansas, Georgia, Mississippi, Oklahoma, North Carolina, South Carolina, Tennessee, Texas and Puerto Rico. They have even made their way to California, Missouri and across the Pacific to Hawaii.⏎The insidious fire ants

  5. #5 peak

    river she notices a white rabbit and follows it into its rabbit hole, emerging again in Wonderland, a fantastical place where meeting mad characters is easy but finding your way home all the while keeping your sanity is less so.⏎Carroll, Lewis.

  6. #6 peak

    experience. We had to pioneer suspensions of liberty. We had to march through strange swamps of total mobilization of civilian effort. We had to find our way in unknown and ambushed forests of peace-making and through the unrevealed and

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 the issue in the company. I have been thinking about the issue in the company. I have been thinking about the issue in the company. I have been thinking about the issue in the company. **The issue in

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

    " which is used to correct skewness in decoder application.⏎And version 7 or higher has "version information" which has information of version itself.⏎next page : How to encode - Capacity. Encoding to data code words.⏎ next page<|endoftext|>

  2. #2 peak

    , let's understand the abstraction in Java first. A Class can be defined as a template / blueprint for creating objects which defines its state and behavior. variable x: Remember from the Java Syntax chapter that a class should always start with an uppercase first

  3. #3 peak

    5 cm⏎Transmission or reproduction of materials protected by copyright beyond that allowed by fair use requires the written permission⏎of the copyright owners. Works not in the public domain cannot be commercially exploited without permission of the copyright⏎owner. Responsibility for any use

  4. #4 peak

    wind 雨 : rain 声 : sound⏎Meaning for Line 4 花

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. L0:1224 suppresses
  2. L0:178 suppresses
  3. L0:1301 weakly suppresses
  4. L1:1853 weakly suppresses
  5. L1:810 weakly suppresses
  6. L1:2509 weakly suppresses
  7. L1:2922 weakly excites
  8. L1:915 weakly suppresses
  9. L1:2444 weakly suppresses
  10. L1:1978 weakly suppresses
  11. L1:332 weakly excites
  12. L1:628 weakly suppresses
  13. L1:2343 weakly excites
  14. L0:157 weakly suppresses
  15. L0:944 weakly excites
  16. L1:501 weakly suppresses
  17. L1:1340 weakly suppresses
  18. L0:2066 weakly suppresses
  19. L1:2264 weakly excites
  20. L1:2801 weakly excites
  21. L1:2208 weakly suppresses
  22. L1:1013 weakly excites
  23. L0:3050 weakly excites
  24. L1:603 weakly excites

Feeds later layers

  1. L21:521 suppresses
  2. L16:1023 suppresses
  3. L18:363 suppresses
  4. L17:1224 suppresses
  5. L4:557 weakly excites
  6. L20:2746 weakly excites
  7. L9:13 weakly excites
  8. L4:1484 weakly suppresses
  9. L17:1078 weakly excites
  10. L12:1259 weakly suppresses
  11. L25:817 weakly suppresses
  12. L17:678 weakly excites
  13. L16:2804 weakly excites
  14. L3:46 weakly suppresses
  15. L6:2675 weakly suppresses
  16. L19:2308 weakly excites
  17. L5:1009 weakly excites
  18. L23:1548 weakly suppresses
  19. L18:1477 weakly suppresses
  20. L18:670 weakly excites
  21. L27:670 weakly excites
  22. L9:413 weakly suppresses
  23. L19:853 weakly excites
  24. L13:3058 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 -0.46 and -0.441 values between -0.40 and -0.383 values between -0.38 and -0.361 values between -0.36 and -0.344 values between -0.34 and -0.326 values between -0.32 and -0.307 values between -0.30 and -0.284 values between -0.28 and -0.2612 values between -0.26 and -0.2418 values between -0.24 and -0.2233 values between -0.22 and -0.2043 values between -0.20 and -0.1846 values between -0.18 and -0.1667 values between -0.16 and -0.14106 values between -0.14 and -0.12145 values between -0.12 and -0.10225 values between -0.10 and -0.09281 values between -0.09 and -0.07330 values between -0.07 and -0.05415 values between -0.05 and -0.03516 values between -0.03 and -0.01488 values between -0.01 and 0.01360 values between 0.01 and 0.03277 values between 0.03 and 0.05171 values between 0.05 and 0.07107 values between 0.07 and 0.0982 values between 0.09 and 0.1148 values between 0.11 and 0.1336 values between 0.13 and 0.1519 values between 0.15 and 0.1719 values between 0.17 and 0.1915 values between 0.19 and 0.217 values between 0.21 and 0.234 values between 0.23 and 0.253 values between 0.25 and 0.271 values between 0.27 and 0.292 values between 0.29 and 0.315 values between 0.31 and 0.33offnegativepositiveoften

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. 拭
  2. OrDefault
  3. inality
  4. requently
  5. Wis
  6. fists
  7. arson
  8. icity
  9. 铛
  10. 猩

Suppresses

  1. .poly
  2. lot
  3. од
  4. GLOBALS
  5. ------------
  6. dehyde
  7. Egyptian
  8. den
  9. <dd
  10. undert