Neuron Explorer

Qwen3-0.6B

opening parenthesis for a gloss or abbreviation

A neuron in layer 13 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 bracket that comes right after a term to explain it, such as a translation or the full form of an abbreviation. Its negative side fires on 'which' after a comma. Turning it up did not produce any clear related text.

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

    portrayed as being conducted by two dare-devil aviators. Nothing could be further from the truth. In fact Alcock was trained and flew in the RNAS (Royal Naval Air Service) and Brown was trained and flew in the RFC (Royal

  2. #2 peak

    Yoga is a form of exercise that focuses on strength, flexibility and breathing to boost physical and mental wellbeing. The main components of a yoga class are asanas (postures) and movements designed to increase strength and flexibility. Breathing is actively

  3. #3 peak

    .⏎Additional diagnostic tests include:⏎Specific treatment for osteogenesis imperfecta (OI) will be determined by your child's physician based on the following:⏎To

  4. #4 peak

    that you may have gingivitis, schedule an appointment at Something To Smile About for a dental exam.⏎What is gingivitis?⏎Gingivitis is the first and only stage that periodontal disease can be reversed. This is because

  5. #5 peak

    Rates for the Global Poor,"⏎American Economic Journal: Applied Economics,⏎American Economic Association, vol. 3(2), pages 137-66, April.⏎- Angus Deaton & Olivier Dupriez, 200

  6. #6 peak

    both the beginning of the Jewish high holidays and the civil year on the Jewish calendar. It’s a good time to wish your Jewish friends leshana tova (happy New Year), to send greeting cards, and perhaps plant trees in Israel in

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: The it's a problem of the non-constant and the problem of the non-constant, which is the two separate problems of the non-constant and the two separate problems that the are the non-constant, which is the two separate

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

    much, start with five or 10 minutes and gradually build up, adding a minute at a time.⏎Exercise can help keep you stay fit during pregnancy, which will be helpful during labor, and may help limit gestational diabetes.⏎For more

  2. #2 peak

    later. It is possible to send and receive bitcoins using a Bitcoin wallet.⏎Even when the value of an asset drops, it’s possible to make a profit, which is counterintuitive

  3. #3 peak

    help increase flexibility, which is essential for overall health. Improved flexibility can help you perform everyday activities more easily

  4. #4 peak

    continuous running and kicking. Physical contact is against the rules, which makes it a team sport that is less injury-prone than what we know as

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. L12:2704 suppresses
  2. L12:2424 suppresses
  3. L6:904 suppresses
  4. L7:1253 suppresses
  5. L12:1095 excites
  6. L10:271 excites
  7. L10:206 excites colon or equals sign before a definition
  8. L5:1035 weakly excites
  9. L11:1910 weakly suppresses
  10. L9:1065 weakly suppresses
  11. L9:298 weakly suppresses
  12. L7:2850 weakly excites
  13. L10:2003 weakly excites
  14. L11:2066 weakly suppresses
  15. L12:2240 weakly suppresses
  16. L8:485 weakly suppresses
  17. L6:1551 weakly suppresses
  18. L6:1620 weakly excites 'in other words' restatement cues
  19. L7:2184 weakly suppresses
  20. L8:1885 weakly suppresses
  21. L5:952 weakly suppresses
  22. L6:1324 weakly excites
  23. L6:1471 weakly excites
  24. L12:2606 weakly excites

Feeds later layers

  1. L16:637 excites
  2. L23:1100 suppresses
  3. L19:3020 suppresses
  4. L20:930 suppresses
  5. L16:918 suppresses
  6. L14:1592 weakly suppresses
  7. L14:1244 weakly suppresses
  8. L15:2030 weakly suppresses
  9. L15:2554 weakly suppresses
  10. L15:2092 weakly suppresses
  11. L15:2938 weakly suppresses comma after an if-clause
  12. L14:1538 weakly suppresses
  13. L21:2380 weakly suppresses
  14. L18:2246 weakly suppresses
  15. L15:1543 weakly suppresses
  16. L22:1253 weakly suppresses
  17. L16:2177 weakly suppresses
  18. L17:984 weakly suppresses
  19. L22:1807 weakly suppresses
  20. L23:1671 weakly suppresses
  21. L20:984 weakly suppresses
  22. L15:1875 weakly excites
  23. L17:1508 weakly suppresses
  24. L16:2609 weakly suppresses

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 -2.38 and -2.271 values between -2.27 and -2.151 values between -1.92 and -1.803 values between -1.80 and -1.681 values between -1.68 and -1.571 values between -1.57 and -1.452 values between -1.45 and -1.331 values between -1.21 and -1.101 values between -1.10 and -0.983 values between -0.98 and -0.8610 values between -0.86 and -0.7510 values between -0.75 and -0.6319 values between -0.63 and -0.5134 values between -0.51 and -0.4076 values between -0.40 and -0.28269 values between -0.28 and -0.16796 values between -0.16 and -0.051805 values between -0.05 and 0.07610 values between 0.07 and 0.19179 values between 0.19 and 0.3154 values between 0.31 and 0.4213 values between 0.42 and 0.548 values between 0.54 and 0.663 values between 0.66 and 0.771 values between 0.77 and 0.891 values between 0.89 and 1.011 values between 1.12 and 1.241 values between 1.24 and 1.362 values between 1.59 and 1.711 values between 2.17 and 2.29offnegativepositiveoften

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. abol
  2. óng
  3. ện
  4. useMemo
  5. 宪
  6. ähr
  7. theid
  8. รณ
  9. @param
  10. ув

Suppresses

  1. ilee
  2. iped
  3. 跡
  4. ethyst
  5. rophic
  6. ersh
  7. التجارية
  8. anical
  9. iy
  10. ASIC