Neuron Explorer

Qwen3-0.6B

unclear: words like 'different', 'close', 'all'

A neuron in layer 22 of Qwen3-0.6B. Its strongest snippets are fairly consistent: several of them peak on the same token. It is featured, so it has steering outputs.

What this neuron does

This neuron fires on a mixed set of ordinary words, such as 'different', 'close', 'all' and 'agree', that tend to sit before a qualifier like 'levels' or 'walks of life'. It does not show one clear topic, and turning it up made the model drift into unrelated text rather than anything tied to these words.

Written by Claude from this neuron's activations and steering outputs. Confidence: low. 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

    ke?⏎Even players with the same language background and goals can play together. In other words, though both my daughter and I desire to learn Norwegian and are at different levels ourselves, we can successfully play the game together and learn from one another.

  2. #2 peak

    we think this problem is relevant for Professor Finnegan's class at WSU.⏎What textbook is this problem found in?⏎Our data indicates that this problem or a close variation was asked in Chemistry: An Atoms First Approach - Zumdahl At

  3. #3 peak

    monastery in sixth-century Egypt is impossible, the lives of late antique Roman elites and Egyptian monks did intersect. Desert Fathers, their monks, and their visitors came from all walks of life and practiced asceticism with varying degrees of rigor. Certainly,

  4. #4 peak

    0 states with the biggest one-year increases in CO2 emissions, with Georgia, Arizona, California, Pennsylvania, Michigan, Iowa, Illinois, Virginia and North Carolina close behind.⏎The top three states – Texas, Georgia and Arizona – had the greatest

  5. #5 peak

    gleaming, cheap , efficient and connects delhi end to end along with its neighboring cities.⏎Delhi has been home to many people for centuries and people from all part of India have stayed here and have enriched the city with their ethos and customs

  6. #6 peak

    ?⏎In order for a sentence to be grammatically correct, the subject and verb must both be singular or plural. In other words, the subject and verb must agree with one another in their tense. If the subject is in plural form, 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 know that, on any two any two 2008 2008 2008 2008 2008 2008 2008 200

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

    , e-mail FormulateThis@allured.com.⏎Question: I’m trying to understand cosmetic emulsion inversion. What actually happens inside the emulsion at the phase borders?⏎Answer: First, remember the HLB system. For ethoxyl

  2. #2 peak

    As the temperature is lowered, the HLB increases and there is a point at which the emulsifiers then orient to form an o/w emulsion (the Phase Inversion Temperature or PIT), resulting in an inverse emulsion.⏎From a theoretical

  3. #3 peak

    a phase boundary, for instance, gas-oil, oil-water, and<|endoftext|>Green kitchen

  4. #4 peak

    they do add up to equal the total, but a voltmeter doesn't provide phase angle measurements, only amplitude.⏎Try measuring voltage dropped across both resistors at once

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. L18:2885 suppresses
  2. L7:653 weakly suppresses
  3. L19:2520 weakly excites
  4. L8:1337 weakly excites
  5. L20:635 weakly excites
  6. L13:102 weakly excites
  7. L9:1994 weakly excites unclear: sun angles, geometry and road-safety terms
  8. L19:1160 weakly excites
  9. L9:153 weakly excites
  10. L6:398 weakly excites
  11. L8:1098 weakly suppresses
  12. L3:1524 weakly excites
  13. L6:2162 weakly suppresses
  14. L19:2904 weakly excites
  15. L8:2996 weakly suppresses
  16. L0:2694 weakly suppresses
  17. L21:149 weakly excites
  18. L8:689 weakly excites
  19. L3:208 weakly suppresses
  20. L16:2314 weakly excites
  21. L10:1553 weakly excites
  22. L14:2550 weakly excites
  23. L21:1759 weakly suppresses
  24. L10:558 weakly excites

Feeds later layers

  1. L24:423 excites
  2. L23:2358 weakly excites
  3. L27:1164 weakly excites
  4. L27:1525 weakly excites
  5. L23:1527 weakly excites
  6. L25:1612 weakly excites
  7. L27:1704 weakly excites
  8. L24:835 weakly excites
  9. L24:2486 weakly excites
  10. L25:2597 weakly excites
  11. L27:2064 weakly excites
  12. L27:903 weakly excites
  13. L23:1015 weakly suppresses
  14. L26:2513 weakly excites
  15. L24:1942 weakly suppresses
  16. L23:2774 weakly suppresses
  17. L25:1049 weakly excites
  18. L26:842 weakly excites
  19. L25:432 weakly suppresses
  20. L25:2565 weakly excites
  21. L24:1660 weakly excites
  22. L24:1629 weakly excites
  23. L23:989 weakly excites
  24. L27:1844 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 -21.69 and -20.801 values between -20.80 and -19.901 values between -12.77 and -11.882 values between -11.88 and -10.981 values between -10.98 and -10.092 values between -10.09 and -9.207 values between -7.42 and -6.524 values between -6.52 and -5.6311 values between -5.63 and -4.7414 values between -4.74 and -3.8512 values between -3.85 and -2.9629 values between -2.96 and -2.0674 values between -2.06 and -1.17569 values between -1.17 and -0.282640 values between -0.28 and 0.61329 values between 0.61 and 1.50103 values between 1.50 and 2.4044 values between 2.40 and 3.2924 values between 3.29 and 4.1813 values between 4.18 and 5.075 values between 5.07 and 5.9610 values between 5.96 and 6.865 values between 6.86 and 7.751 values between 7.75 and 8.642 values between 8.64 and 9.531 values between 10.42 and 11.322 values between 11.32 and 12.211 values between 13.10 and 13.99offnegativepositiveoften

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. 物理
  2. sociale
  3. soci
  4. phys
  5. sociology
  6. sociales
  7. 哲学
  8. geography
  9. socially
  10. 地理

Suppresses

  1. LoginForm
  2. pez
  3. 某
  4. 贴
  5. Luxembourg
  6. .lazy
  7. licit
  8. lik
  9. 查看全文
  10. ılan