Neuron Explorer

Qwen3-0.6B

comma after a speech verb

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 fires on a comma that comes right before a quotation, such as "He says," or "You may be asking yourself,". When it was turned up, the writing did not clearly show this habit.

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

    (OLST)1,2 is a simple, easy and effective method to screen for balance impairments in the older adult population.⏎You may be asking yourself, "how can standing on one leg provide you with any information about balance, after

  2. #2 peak

    .⏎What’s the Best Humidity Level for your Home? How do you maintain the level throughout the year?⏎You may have thought, “What is the ideal level for my house in the summer or winter months?”

  3. #3 peak

    watches Gertrude's reaction to the statement. By her reaction, he may be able to judge whether she has knowledge of Claudius' crime. He says, "Almost as bad...As kill a king, and marry with his brother."

  4. #4 peak

    isticexpialidocious?"⏎- Our speaker, sensing our sensory overload, steers us back on course. At the end of line 9, he says, "the gaps I mean," reminding us of the matter at hand (the broken

  5. #5 peak

    ).⏎It is interesting to note, however, that while the Torah commands us to verbally recall the Exodus from Egypt twice

  6. #6 peak

    to gauge.⏎Recent technological and scientific advances, particularly the recovery and sequencing of ancient DNA from old bones, are revealing some of that lost history.⏎We now suspect, for example, that the bacterium has been infecting people since at least the

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 working on the idea of a "data" and "data" is the same as a "data" and "data" is the same as a "data" and "data" is the same as a "data"

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

    indirect techniques only calculate it from other indirect indicators.⏎Image Credit: Soonthorn Wongsaita / Shutterstock⏎Many devices for LOD measurement use the thermogravimetric principle. This uses the total loss of weight incurred by the sample on

  2. #2 peak

    He professes to know how teachers should teach in the entire country and what they should teach, and yet, he is not a teacher. Coleman worked for McGraw-Hill (textbook publishing company) before leaving to create GROW a curriculum

  3. #3 peak

    of the cathedral in the Romanesque or Norman style, starting with the nave. Robert died shortly afterward, but the work was carried forward by Geoffrey de Montbray, his successor as bishop, appointed in 1048. Geoffrey

  4. #4 peak

    at bottom left showing two islands discovered by Cook and Marion du Fresne. One of the few maps in the volume drawn by Bonne's son. Engraved by Andre. Condition: A nice impression with minor offsetting and a few faint

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. L0:714 weakly excites
  2. L0:696 weakly suppresses
  3. L1:126 weakly excites
  4. L0:729 weakly suppresses 'st' in dates like 'June 1st'
  5. L1:2796 weakly excites
  6. L1:2342 weakly excites
  7. L1:1125 weakly suppresses
  8. L1:1467 weakly suppresses
  9. L0:2968 weakly suppresses
  10. L1:2570 weakly suppresses
  11. L0:251 weakly suppresses
  12. L1:2736 weakly excites
  13. L0:173 weakly suppresses
  14. L1:915 weakly suppresses
  15. L0:263 weakly suppresses
  16. L1:909 weakly excites
  17. L0:1472 weakly excites
  18. L0:1524 weakly suppresses
  19. L1:2119 weakly excites
  20. L0:2204 weakly suppresses
  21. L0:2774 weakly excites
  22. L1:384 weakly suppresses
  23. L1:445 weakly excites document start marker token
  24. L1:507 weakly excites

Feeds later layers

  1. L7:1833 excites
  2. L6:700 weakly excites
  3. L6:600 weakly excites
  4. L11:2311 weakly excites
  5. L4:622 weakly suppresses
  6. L4:2378 weakly excites
  7. L16:3066 weakly excites
  8. L18:647 weakly suppresses
  9. L8:2936 weakly excites
  10. L3:1537 weakly suppresses
  11. L8:1105 weakly excites
  12. L4:2605 weakly suppresses
  13. L25:860 weakly excites
  14. L11:2007 weakly suppresses
  15. L10:2026 weakly excites
  16. L8:1312 weakly excites
  17. L20:2412 weakly suppresses
  18. L4:679 weakly excites
  19. L7:1081 weakly excites
  20. L8:2957 weakly excites
  21. L15:1745 weakly suppresses
  22. L13:1145 weakly excites
  23. L13:1302 weakly suppresses
  24. L16:2796 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 -0.83 and -0.681 values between -0.68 and -0.549 values between -0.54 and -0.3951 values between -0.39 and -0.24464 values between -0.24 and -0.092121 values between -0.09 and 0.061103 values between 0.06 and 0.20123 values between 0.20 and 0.3513 values between 0.35 and 0.509 values between 0.50 and 0.653 values between 0.65 and 0.803 values between 0.80 and 0.941 values between 0.94 and 1.093 values between 1.24 and 1.391 values between 2.27 and 2.421 values between 4.49 and 4.641 values between 4.93 and 5.08offnegativepositiveoften

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. enes
  2. leen
  3. 切尔
  4. adamente
  5. enet
  6. ewise
  7. 人心
  8. лим
  9. GenericType
  10. contracting

Suppresses

  1. avery
  2. łów
  3. trạng
  4. ớt
  5. hog
  6. itution
  7. itive
  8. berg
  9. 是一
  10. ición