Neuron Explorer

Qwen3-0.6B

conservation laws and invariance

A neuron in layer 4 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 responds to physics wording about things that stay the same, such as "conservation", "invariant" and "conserved". Turning it up pushed the text toward maths and physics ideas, though not always coherently.

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

    . An invariant profile was observed when movements were normalized for velocity and distance. Two arms, extended

  2. #2 peak

    Noether's theorem, which states that there is a one-to-one correspondence between each one of them and a differentiable symmetry of nature. For example, the conservation of energy follows from the time-invariance of physical systems, and the conservation

  3. #3 peak

    to symmetries and invariance principles, three special conservation laws have been described, associated with inversion or reversal of space, time, and charge.⏎Conservation laws are considered to be fundamental laws of nature, with broad application in physics,

  4. #4 peak

    , which apply to such quantities as mass, parity, lepton number, baryon number, strangeness, hypercharge, etc. These quantities are conserved in certain classes of physics processes, but not in all.⏎A local conservation law

  5. #5 peak

    negative numbers⏎<|endoftext|>In physics, a conservation law states that a particular measurable property of an isolated physical system does not change as the system evolves over time. Exact conservation laws include conservation of energy, conservation of linear momentum, conservation of angular momentum,

  6. #6 peak

    change form. In general, the total quantity of the property governed by that law remains unchanged during physical processes. With respect to classical physics, conservation laws include conservation of energy, mass (or matter), linear momentum, angular momentum, and electric charge

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: (A) A (B) A (C) A (D) A Let me think about this. The question is about a certain item. Which of the following is the correct one? A. A

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

    involvement.⏎Introduction. Part 1. Geometry and Nature. Chaos game visualization of sequences (H. J. Jeffrey). Tumor growth simulation (W

  2. #2 peak

    equal.⏎Label this plum tree graph to make it totally magic!⏎A graph is a set of vertices (or nodes), together with a set of edges (or arcs).⏎Look at

  3. #3 peak

    ⏎From symmetric meson theory, the formation of helium nucleus from two deuterons, i.e., fusion, requires necessarily a direct force due to exchange of

  4. #4 peak

    the⏎major arc AB,⏎where you go the long way around the bottom of the circle. Unless stated otherwise, it always means the⏎minor arc - the shortest of the two.⏎If you want to indicate the major arc

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. L3:105 excites
  2. L3:25 suppresses
  3. L3:798 suppresses
  4. L3:50 excites
  5. L0:178 suppresses
  6. L0:1224 suppresses
  7. L3:52 weakly suppresses
  8. L3:91 weakly excites
  9. L0:1301 weakly suppresses
  10. L3:448 weakly suppresses
  11. L3:1990 weakly suppresses
  12. L2:1899 weakly excites
  13. L1:1464 weakly excites
  14. L3:556 weakly suppresses
  15. L3:421 weakly suppresses
  16. L3:2850 weakly suppresses
  17. L3:115 weakly suppresses
  18. L2:1809 weakly excites
  19. L2:2349 weakly suppresses word pieces inside cinnamon, China, cabar
  20. L3:102 weakly suppresses
  21. L3:185 weakly suppresses
  22. L2:1334 weakly suppresses
  23. L2:160 weakly suppresses
  24. L2:2759 weakly excites

Feeds later layers

  1. L7:2422 excites
  2. L10:1001 weakly excites
  3. L12:786 weakly excites
  4. L8:1934 weakly excites
  5. L17:2082 weakly suppresses
  6. L9:477 weakly excites
  7. L7:968 weakly excites
  8. L5:2894 weakly excites
  9. L9:417 weakly excites
  10. L11:2354 weakly excites
  11. L13:1246 weakly suppresses
  12. L9:1673 weakly excites
  13. L11:2307 weakly excites
  14. L16:1732 weakly excites
  15. L12:1259 weakly suppresses
  16. L8:912 weakly excites
  17. L16:2051 weakly suppresses
  18. L21:659 weakly suppresses end of document boundary
  19. L9:1959 weakly excites
  20. L23:875 weakly suppresses
  21. L5:1881 weakly excites
  22. L11:2468 weakly excites
  23. L10:2774 weakly excites
  24. L17:1005 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 -0.60 and -0.562 values between -0.45 and -0.416 values between -0.41 and -0.3713 values between -0.37 and -0.3318 values between -0.33 and -0.2923 values between -0.29 and -0.2564 values between -0.25 and -0.22105 values between -0.22 and -0.18161 values between -0.18 and -0.14233 values between -0.14 and -0.10415 values between -0.10 and -0.06532 values between -0.06 and -0.02691 values between -0.02 and 0.01518 values between 0.01 and 0.05396 values between 0.05 and 0.09283 values between 0.09 and 0.13167 values between 0.13 and 0.17115 values between 0.17 and 0.2171 values between 0.21 and 0.2446 values between 0.24 and 0.2820 values between 0.28 and 0.3215 values between 0.32 and 0.365 values between 0.36 and 0.403 values between 0.40 and 0.443 values between 0.44 and 0.471 values between 0.47 and 0.511 values between 0.90 and 0.94offnegativepositiveoften

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. OLLOW
  2. nants
  3. 功率
  4. давать
  5. inciple
  6. 틴
  7. iterate
  8. 纲
  9. transit
  10. →super

Suppresses

  1. istical
  2. quirer
  3. mock
  4. opard
  5. zzle
  6. isnan
  7. OMET
  8. ostringstream
  9. azio
  10. izzo