Neuron Explorer

Qwen3-0.6B

"g" in "e.g." examples

A neuron in layer 1 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 the "g" in "e.g." when it introduces examples in brackets, such as "(e.g. eagle shrieks)". Turning it up or down gave replies on unrelated subjects, so the steering results do not show what it does.

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

    , salinity) affected the physiology, reproductive capacity, and survivorship of marine organisms. Likewise, Comparative Physiology seeks to unify understanding of common physiological processes (e.g., osmoregulation, respiration) across many taxa; in this un

  2. #2 peak

    and by eagles and leopards in the Tai Forest.⏎Using transcriptions of these monkey calls gathered in field experiments involving playbacks of predator calls (e.g. eagle shrieks and leopard growls), the researchers found greater complexity in expression

  3. #3 peak

    and perhaps most importantly,⏎conduct research on those behaviors with the greatest environmental impact (e.g., transportation choices, particularly in the U.S.) rather than those that are more

  4. #4 peak

    not advisable. Additionally, homebound instruction may be used to supplement the classroom program for health-impaired children whose conditions may interfere with consistent school attendance, (e.g., students receiving dialysis, radiation or chemotherapy). The inability to attend school must

  5. #5 peak

    the solid Earth is composed, the processes that generate Earth's landforms, natural hazards associated with geologic processes, geologic time, and surface processes (e.g., glaciers

  6. #6 peak

    the performance of methods designed to eliminate edge-effect bias. We show that the reflection method solves the edge-problem for designs that use symmetric radial transects (e.g., straight-line and X-shaped transects centered at the sample point). This method

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 am not sure about this, but I do like the rest of the week. But I have been thinking about lately: "I am not sure about this, but I do like the rest of the week." But I have been

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

    (i) Khadi cloth was often more expensive than mass produced mill cloth and poor people

  2. #2 peak

    (A) tails. This prevents oligo d(T) mediated mRNA isolation. Here,

  3. #3 peak

    (2) failing to provide food, care or shelter;⏎(3) abandoning an

  4. #4 peak

    Currents⏎An alternating voltage may be generated⏎(1) by rotating a coil at constant angular velocity in a uniform magnetic field as shown in Fig. (a).⏎(2) by rotating a magnetic field at a constant angular velocity within a

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 suppresses
  4. L0:1 weakly suppresses
  5. L0:13 weakly excites
  6. L0:595 weakly excites
  7. L0:2284 weakly excites
  8. L0:1404 weakly excites
  9. L0:16 weakly excites
  10. L0:1672 weakly suppresses
  11. L0:1062 weakly excites
  12. L0:1619 weakly suppresses
  13. L0:2362 weakly suppresses
  14. L0:1991 weakly excites
  15. L0:489 weakly excites
  16. L0:1705 weakly suppresses
  17. L0:210 weakly suppresses
  18. L0:2004 weakly excites
  19. L0:564 weakly suppresses
  20. L0:2184 weakly excites
  21. L0:1060 weakly suppresses
  22. L0:1249 weakly suppresses
  23. L0:1935 weakly excites
  24. L0:581 weakly excites

Feeds later layers

  1. L4:45 excites
  2. L3:1111 weakly excites
  3. L19:1797 weakly suppresses
  4. L19:1966 weakly excites
  5. L12:1146 weakly excites "but" after a concession
  6. L15:628 weakly excites
  7. L25:2822 weakly excites
  8. L5:1999 weakly excites
  9. L20:2253 weakly suppresses
  10. L11:2058 weakly suppresses
  11. L18:1090 weakly excites unclear: mid-sentence words before adverbs
  12. L22:2677 weakly excites
  13. L18:2938 weakly suppresses
  14. L8:1998 weakly suppresses
  15. L27:1565 weakly excites
  16. L20:104 weakly suppresses
  17. L27:548 weakly suppresses
  18. L15:3009 weakly suppresses
  19. L18:620 weakly suppresses
  20. L20:2344 weakly suppresses
  21. L3:1329 weakly excites
  22. L20:1743 weakly excites
  23. L5:1556 weakly suppresses
  24. L24:1527 weakly excites space before a number, often an age

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.

7 values between -0.44 and -0.3150 values between -0.31 and -0.181082 values between -0.18 and -0.052511 values between -0.05 and 0.08235 values between 0.08 and 0.2119 values between 0.21 and 0.351 values between 0.35 and 0.481 values between 1.53 and 1.661 values between 1.66 and 1.791 values between 4.69 and 4.82offnegativepositiveoften

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. ellan
  2. fühl
  3. рен
  4. ул
  5. reon
  6. iron
  7. icana
  8. at
  9. 饫
  10. .NoError

Suppresses

  1. đích
  2. tracts
  3. getStore
  4. ög
  5. 目的
  6. .hxx
  7. _CTX
  8. colours
  9. prizes
  10. (co