Neuron Explorer

Qwen3-0.6B

period inside abbreviations like i.e. and e.g

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 dot in shorthand like "i.e." or "e.g.", the dot that sits in the middle of an abbreviation rather than ending a sentence. It went negative on dots that close off a sentence. Turning it up made the model's text drift into unrelated topics, so the steering results don't show a clear effect.

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

    Gemeente, 1883-1884: v. I, Nos. 1571-1581; v. II, 2, Nos. 703-739.⏎

  2. #2 peak

    Have a snack supply for the day; e.g. mandarins, apples, bananas, low fat yoghurts, dried

  3. #3 peak

    to the small population numbers limiting the number of mutational possibilities; and the much larger genome, which cannot sustain a similar mutation rate without error catastrophe; i.e. extinction; and sexual reproduction means that there is 50%

  4. #4 peak

    et al. 2020). Under these circumstances, tree´s survival primarily depends on the extent to which tree function is impaired by drought (i.e. resistance, Lloret et al., 2011). To increase

  5. #5 peak

    injection molding processes, as well as post-production and finishing procedures, to predict how critical process parameters (CPPs), e.g. melt temperature or injection velocity, would affect real time product quality, durability and reliability

  6. #6 peak

    zach-Donau region. Technically, the metadata database is a graph database. This technology aims at semantically linking data to each other, i.e. meaningfully. In concrete terms, this means that information is linked in a triple

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 lately. I have been thinking about lately. I have been thinking about lately. I have been thinking about lately. I have been thinking about lately. I have been thinking about lately. I have been thinking about lately

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

    instills a sense of responsibility and⏎achievement, while furthering a child's desire to learn. This interactive⏎activity is also great for children of all ages.⏎The benefits of

  2. #2 peak

    high resistance to rust,and high fabrication ability.Cost-effective and versatile material,304 sheet is suitable for most applications where durable support and corrosion resistance are required.Stainless steel sheet metal and plate also provides excellent weight<|endoftext|>US Pharm. 2

  3. #3 peak

    as the corpus), and⏎the actual practice (conducted in the arena).⏎These two continually learn from each other and evolve.⏎The beginnings of the science⏎

  4. #4 peak

    the majority of the archeological material is within the top ten⏎centimeters of the soil. The damage from the proposed construction should⏎be considered mitigated.⏎Site 48TE414 is in the South Jenny Lake development.

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. L0:1301 suppresses
  3. L0:178 suppresses
  4. L0:348 weakly excites space token right after a period
  5. L0:36 weakly excites
  6. L0:51 weakly excites
  7. L0:0 weakly excites
  8. L0:447 weakly suppresses
  9. L0:799 weakly suppresses
  10. L0:1001 weakly suppresses
  11. L0:2113 weakly suppresses
  12. L0:2838 weakly suppresses
  13. L0:835 weakly excites
  14. L0:212 weakly suppresses
  15. L0:2137 weakly suppresses
  16. L0:1567 weakly excites
  17. L0:2830 weakly excites
  18. L0:560 weakly excites
  19. L0:2257 weakly suppresses
  20. L0:990 weakly excites
  21. L0:749 weakly suppresses
  22. L0:2376 weakly excites
  23. L0:1186 weakly excites
  24. L0:2791 weakly suppresses

Feeds later layers

  1. L17:3027 weakly excites
  2. L26:380 weakly excites
  3. L2:694 weakly excites
  4. L22:742 weakly excites
  5. L21:524 weakly suppresses
  6. L19:328 weakly excites
  7. L3:259 weakly suppresses
  8. L10:213 weakly suppresses
  9. L3:2930 weakly excites
  10. L17:949 weakly suppresses
  11. L18:1085 weakly suppresses full stops ending sentences
  12. L15:1998 weakly excites
  13. L8:740 weakly excites
  14. L11:841 weakly excites
  15. L6:62 weakly suppresses
  16. L8:2128 weakly suppresses
  17. L19:2680 weakly excites
  18. L22:481 weakly excites
  19. L6:59 weakly suppresses
  20. L19:630 weakly excites
  21. L22:1596 weakly excites
  22. L4:1470 weakly excites
  23. L15:1334 weakly suppresses
  24. L3:2785 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 -1.23 and -1.171 values between -1.17 and -1.112 values between -0.81 and -0.754 values between -0.70 and -0.645 values between -0.64 and -0.585 values between -0.58 and -0.526 values between -0.52 and -0.4610 values between -0.46 and -0.4014 values between -0.40 and -0.3428 values between -0.34 and -0.2861 values between -0.28 and -0.22126 values between -0.22 and -0.16269 values between -0.16 and -0.10600 values between -0.10 and -0.041019 values between -0.04 and 0.02923 values between 0.02 and 0.08527 values between 0.08 and 0.14195 values between 0.14 and 0.2070 values between 0.20 and 0.2624 values between 0.26 and 0.329 values between 0.32 and 0.371 values between 0.43 and 0.491 values between 0.55 and 0.611 values between 0.61 and 0.671 values between 0.67 and 0.731 values between 0.73 and 0.792 values between 0.79 and 0.851 values between 1.03 and 1.091 values between 1.09 and 1.15offnegativepositiveoften

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. 瓢
  2. openh
  3. 叨
  4. )':
  5. 氡
  6. 醋
  7. 饮
  8. 一杯
  9. מין
  10. Uran

Suppresses

  1. ctors
  2. stances
  3. itives
  4. itive
  5. cư
  6. leine
  7. /lgpl
  8. cba
  9. sumer
  10. successes