Neuron Explorer

Qwen3-0.6B

antibiotic and drug-name context

A neuron in layer 24 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 is active in medical text about antibiotics and bacteria, such as 'Ampicillin is a penicillin antibiotic'. It works against the word 'form', and it goes quiet or negative right before 'isoform' or 'form'. Turning it up made the model write unrelated sentences and turning it down made it repeat 'form'.

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

    extended-spectrum beta-lactamase with additional resistance to carbenicillin, amoxicillin, and increasingly to ceftazidime. The bacteria remain

  2. #2 peak

    in northern Puerto Rico. Particular attention is given to chlorinated volatile organic compounds and phthalates, although biomarkers of phenols, metals, and parabens exposure are also being explored as precursors of preterm birth. Identification of

  3. #3 peak

    floxacin is an antibiotic that is part of the fluoroquinolone family. It effectively kills bacteria, while dexameth

  4. #4 peak

    it is growing. This weakens the cell wall and it ruptures resulting in the death of the bacteria.⏎If you miss a dose of Ampicillin and are taking it regularly, take it as soon as possible. If it

  5. #5 peak

    few days.⏎Ask your health care provider any questions you may have about how to use Ampicillin.⏎Drug Class and Mechanism⏎Ampicillin is a penicillin antibiotic. It works by killing sensitive bacteria by interfering with formation of

  6. #6 peak

    affect people with underlying diseases, such as alcoholism, diabetes and chronic lung disease.⏎Treatment and resistance⏎Treatment for Klebsiella pneumonia is by antibiotics such as aminoglycosides and cephalosporins, the choice depending

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: form form form for form form for form form for form form for form form for form form for form form for form form for form form for form form for form form for form form for form form for form form for form form for form form

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

    no overlap between the two diseases, the researchers noted.⏎One chaperone that protected against ALS aggregates — DNAJB14 — exists in two versions, called isoforms. The isoforms are similar, but one is shorter and lacks some important

  2. #2 peak

    Adrian Krainer. (click to enlarge)⏎One difference between metabolism in cancer and normal cells is the switch in cancer to the production of a different version, or isoform, of a protein produced from the pyruvate kinase-M (PK

  3. #3 peak

    Old French practiser or medieval Latin practizare, alteration of practicare ‘perform, perform’, from practica ‘apply’, from Greek praktikē, feminine (

  4. #4 peak

    While there are less severe and later-onset versions of citrullinemia type I, the mutations for which Counsyl screens are associated with the more severe form that affects infants shortly after birth. It is also known as “classic”

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. L22:331 suppresses
  2. L19:2275 excites
  3. L8:1575 weakly excites
  4. L9:228 weakly suppresses
  5. L19:1251 weakly excites
  6. L6:2734 weakly excites
  7. L19:2469 weakly suppresses
  8. L20:1497 weakly excites
  9. L3:2101 weakly suppresses
  10. L8:611 weakly excites
  11. L10:2883 weakly suppresses
  12. L23:706 weakly excites
  13. L3:37 weakly suppresses
  14. L3:1029 weakly suppresses
  15. L11:2297 weakly excites
  16. L7:1778 weakly excites
  17. L15:3057 weakly excites
  18. L10:2954 weakly excites
  19. L6:2383 weakly suppresses
  20. L1:3039 weakly excites
  21. L6:1095 weakly excites
  22. L22:172 weakly excites
  23. L0:1422 weakly suppresses
  24. L0:2043 weakly excites

Feeds later layers

  1. L27:966 strongly suppresses
  2. L26:1710 suppresses
  3. L25:1400 suppresses
  4. L26:1407 suppresses
  5. L27:2151 weakly suppresses
  6. L25:2740 weakly excites
  7. L26:435 weakly excites
  8. L25:618 weakly suppresses
  9. L25:2710 weakly excites
  10. L25:1388 weakly suppresses
  11. L25:928 weakly suppresses
  12. L25:1170 weakly excites
  13. L25:62 weakly excites
  14. L26:349 weakly suppresses
  15. L26:2446 weakly excites
  16. L27:1535 weakly suppresses
  17. L26:1154 weakly excites
  18. L26:1254 weakly suppresses
  19. L26:223 weakly excites
  20. L26:2023 weakly excites
  21. L25:504 weakly suppresses
  22. L25:1191 weakly excites
  23. L25:1037 weakly suppresses
  24. L27:1037 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 -31.31 and -29.921 values between -16.00 and -14.612 values between -14.61 and -13.214 values between -13.21 and -11.826 values between -11.82 and -10.433 values between -10.43 and -9.044 values between -9.04 and -7.6511 values between -7.65 and -6.2521 values between -6.25 and -4.8632 values between -4.86 and -3.4781 values between -3.47 and -2.08234 values between -2.08 and -0.683277 values between -0.68 and 0.71187 values between 0.71 and 2.1015 values between 2.10 and 3.498 values between 3.49 and 4.889 values between 4.88 and 6.283 values between 6.28 and 7.671 values between 7.67 and 9.062 values between 9.06 and 10.451 values between 10.45 and 11.851 values between 11.85 and 13.241 values between 13.24 and 14.631 values between 14.63 and 16.021 values between 16.02 and 17.411 values between 22.98 and 24.38offnegativepositiveoften

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. abay
  2. vements
  3. redits
  4. 委宣传
  5. pery
  6. icken
  7. ثقة
  8. aurants
  9. ouch
  10. lâ

Suppresses

  1. form
  2. forms
  3. form
  4. Forms
  5. Form
  6. Form
  7. forms
  8. _form
  9. →form
  10. _forms