Neuron Explorer

Qwen3-0.6B

"not" or "instead" contrasting a correction

A neuron in layer 6 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 words like "not" when they add a correction or contrast, such as "not even" or "not X, not Y". It switches off on "instead" and "but" in contrast phrases. Turning it up gave repetitive or unrelated text, so the steering doesn't 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

    research project at MIT called Sparsh is aiming to fix that oversight.⏎Sparsh (the Hindi word for "touch") isn't an app, at least not in the way we generally use the word. It's a tool that's supposed

  2. #2 peak

    Waziristan. Consequently, it became a sanctuary for fundamentalist terrorists who hold outdated ultraconservative views that are particularly detrimental for women.⏎But nothing — not even a conservative culture or fear of the Taliban<|endoftext|>This preview shows pages 1

  3. #3 peak

    , but the more pressure in the game, the more he seemed to “forget” how to throw.⏎Knoblauch is not the first and probably not the last player this has happened to. Tennis players ‘forget’ how to toss

  4. #4 peak

    them to fulfil their promises no matter what. We are also given access to social security and food parcels from the government so we wouldn’t starve, at least not starve enough to find a job. Those that manage to get a job,

  5. #5 peak

    , it is in a circle, at most perhaps in a widening circle. Man today is not wiser than the ancient seers and sages and thinkers, not more spiritual than the great seekers of old, the first mighty mystics, not

  6. #6 peak

    Rabbi Emil Fackenheim in his book, "God's Presence in History." "Nevertheless, not messengers, not angels, not intermediaries, but God himself acts in human history."⏎God hears

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: what is the value of the question is to be given in a diagram of a cube (a regular hexagon) and a rectangle, and a square, and a circle, and a square of a circle, and a square of a circle

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

    is a semiconductor device that can be programmed to perform various logic functions. Unlike a microcontroller, an FPGA is not designed to execute a specific set of instructions. Instead, developers program the FPGA to perform specific logic functions by configuring its hardware components.

  2. #2 peak

    example, your tomato ketchup or mayonnaise⏎- Stop using plastic drinking straws and instead try metal, paper, or even wheat straws⏎- Swap tampons and

  3. #3 peak

    says, “This he said, not that he cared for the poor, but because he was a thief, and had the money box; and he used to

  4. #4 peak

    any decision, then you will be well-informed of all the consequences of your decision-making. Critical thinking doesn’t mean that your decision will never be wrong. Instead, it increases your chances of making good decisions. Although if the outcome is not

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. L5:45 weakly suppresses
  2. L5:2082 weakly suppresses
  3. L3:807 weakly suppresses comma inside dates and years
  4. L5:57 weakly suppresses
  5. L1:2539 weakly suppresses
  6. L5:60 weakly excites
  7. L1:2028 weakly suppresses
  8. L3:1468 weakly excites
  9. L3:1377 weakly suppresses
  10. L0:2285 weakly excites
  11. L4:248 weakly suppresses
  12. L5:96 weakly suppresses
  13. L2:2027 weakly suppresses
  14. L3:1382 weakly suppresses
  15. L3:964 weakly excites
  16. L3:2126 weakly excites
  17. L2:851 weakly suppresses
  18. L4:1901 weakly suppresses
  19. L0:2946 weakly excites
  20. L5:1326 weakly suppresses
  21. L3:1883 weakly excites
  22. L2:1194 weakly suppresses
  23. L5:169 weakly excites
  24. L1:2681 weakly suppresses

Feeds later layers

  1. L8:1784 suppresses
  2. L8:1876 suppresses
  3. L14:2166 suppresses
  4. L7:678 excites
  5. L8:1011 suppresses
  6. L7:1846 suppresses
  7. L11:1861 suppresses
  8. L12:1071 suppresses
  9. L14:2811 excites
  10. L7:2256 suppresses 'to' after come or bring
  11. L14:2713 excites
  12. L18:1553 suppresses
  13. L9:2343 suppresses
  14. L27:1088 weakly suppresses
  15. L7:1764 weakly excites
  16. L9:1046 weakly suppresses
  17. L13:220 weakly suppresses
  18. L7:2806 weakly excites
  19. L9:1133 weakly suppresses
  20. L9:1472 weakly excites
  21. L11:2662 weakly suppresses
  22. L9:2203 weakly excites
  23. L19:1811 weakly suppresses
  24. L13:92 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.

2 values between -7.79 and -7.491 values between -6.27 and -5.972 values between -2.93 and -2.631 values between -2.63 and -2.322 values between -2.32 and -2.021 values between -1.72 and -1.412 values between -1.41 and -1.114 values between -1.11 and -0.8017 values between -0.80 and -0.50203 values between -0.50 and -0.202715 values between -0.20 and 0.11919 values between 0.11 and 0.4133 values between 0.41 and 0.721 values between 0.72 and 1.021 values between 1.02 and 1.322 values between 1.32 and 1.631 values between 2.23 and 2.541 values between 4.06 and 4.36offnegativepositiveoften

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. Late
  2. alyzed
  3. ��
  4. イベ
  5. 孽
  6. Tucker
  7. ngth
  8. 会有
  9. nier
  10. レビュー

Suppresses

  1. instead
  2. instead
  3. rather
  4. Instead
  5. Instead
  6. .Ptr
  7. 呢
  8. (d
  9. downright
  10. ignum