Neuron Explorer

Qwen3-0.6B

unclear: word before a stated value or definition

A neuron in layer 16 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 hard to pin down. It tends to light up on small words like 'is', ':' or '-' that come right before the point where a value, definition or explanation is about to follow. When it was turned up, the model's text drifted into repeated lists about presidents and cars, which doesn't clearly tell us much.

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

    recognizing alternative interpretations in their writing. (A sample story and four figures of data are included.) (RS)⏎Publication Type: Speeches/Meeting Papers; Reports - Research⏎Education Level: N/A⏎Authoring Institution: N/A<|endoftext|>A

  2. #2 peak

    ). The state of cell is determined by a set of rules, which all cells must obey. These rules are applied to every cell simultaneously. The only player input is the initial world state.⏎Conway's Game of Life is played by creating an

  3. #3 peak

    to other parts of the world to keep it up. Given all these pressures, we should be smart about how we use our resources — and one of our resources is soil.”⏎Bradley is collecting data from researchers who’ve already measured global crop output

  4. #4 peak

    electrical engineering positions of various certification and licensure exams. Examples and case studies are included.⏎File type: Downloadable PDF file (Note: Print book not included. To purchase print copy

  5. #5 peak

    boycott of Uruguay.⏎Format: Knockouts, semi-finals, final⏎Golden boot: Leonidas (Brazil)⏎

  6. #6 peak

    .⏎How Do I Set Up A Website Using Iis?⏎You can access the Internet Information Services (IIS) Manager by selecting Control Panel > Administrative Tools. Select Add Website from the Connections panel, expand your host tree, right-click on Sites

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 to be alone in my work. I have to be alone in my work. I have to be alone in my work. I have to be alone in my work. I have to be alone in my work. I have to

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

    ), association mining (customer analytics), and latent variable models (factor analysis).⏎Semi-supervised machine learning⏎As the name may suggest, this model combines elements of supervised and unsupervised learning. It works by using both labeled and unlabeled

  2. #2 peak

    . Pat skin dry and immediately apply a layer of moisturizer while skin is still slightly damp. Choose one that contains ingredients known to be effective for dry skin such as hyaluronic acid, glycerin, alpha-hydroxy acids or

  3. #3 peak

    steopenia / Osteoporosis⏎Osteopenia is a term used to describe bone density that is somewhat lower than normal, but not low enough to be diagnosed as osteoporosis. A person with osteopenia is

  4. #4 peak

    infrared radiometer indirectly measures the temperature of a very thin layer (about 10 micrometres thick) or skin of the ocean (leading to the phrase skin temperature) representing the top millimeter; a thermometer attached to an already fast

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. L15:2033 suppresses
  2. L10:1748 excites
  3. L15:1438 suppresses
  4. L14:1881 excites
  5. L15:2787 suppresses
  6. L11:1651 suppresses
  7. L15:1588 excites
  8. L13:287 suppresses
  9. L15:1497 suppresses
  10. L15:1176 excites
  11. L14:497 suppresses
  12. L12:2163 excites
  13. L14:2990 excites
  14. L14:2532 weakly excites
  15. L14:2989 weakly suppresses
  16. L14:962 weakly suppresses
  17. L14:376 weakly excites
  18. L14:2904 weakly excites
  19. L14:2235 weakly suppresses
  20. L15:1092 weakly excites
  21. L8:1097 weakly suppresses
  22. L14:2193 weakly suppresses
  23. L14:2244 weakly suppresses
  24. L7:2093 weakly suppresses

Feeds later layers

  1. L17:2057 excites
  2. L17:1151 excites
  3. L17:867 weakly suppresses unclear: punctuation before an explanation
  4. L19:1412 weakly suppresses
  5. L19:666 weakly excites
  6. L18:1088 weakly excites 'and' or 'the' joining two parallel items
  7. L18:989 weakly excites
  8. L23:968 weakly excites
  9. L18:2361 weakly suppresses
  10. L19:924 weakly excites
  11. L18:2204 weakly excites
  12. L19:411 weakly suppresses
  13. L19:354 weakly excites
  14. L18:801 weakly excites
  15. L17:1877 weakly excites
  16. L17:2861 weakly excites
  17. L20:714 weakly suppresses
  18. L22:2649 weakly suppresses
  19. L19:573 weakly excites
  20. L18:2173 weakly excites
  21. L17:2711 weakly excites
  22. L17:2602 weakly excites
  23. L20:731 weakly excites
  24. L17:663 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 -2.68 and -2.511 values between -2.35 and -2.191 values between -2.19 and -2.023 values between -1.86 and -1.691 values between -1.69 and -1.535 values between -1.53 and -1.379 values between -1.37 and -1.208 values between -1.20 and -1.0418 values between -1.04 and -0.8725 values between -0.87 and -0.7157 values between -0.71 and -0.5593 values between -0.55 and -0.38196 values between -0.38 and -0.22597 values between -0.22 and -0.051703 values between -0.05 and 0.11508 values between 0.11 and 0.27243 values between 0.27 and 0.44145 values between 0.44 and 0.6086 values between 0.60 and 0.7756 values between 0.77 and 0.9329 values between 0.93 and 1.0923 values between 1.09 and 1.2629 values between 1.26 and 1.4218 values between 1.42 and 1.5912 values between 1.59 and 1.7513 values between 1.75 and 1.916 values between 1.91 and 2.086 values between 2.08 and 2.243 values between 2.24 and 2.415 values between 2.41 and 2.572 values between 2.73 and 2.902 values between 2.90 and 3.061 values between 3.06 and 3.231 values between 3.23 and 3.392 values between 3.72 and 3.88offnegativepositiveoften

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. 的答案
  2. 短线
  3. bec
  4. 外交
  5. 不过是
  6. FN
  7. became
  8. 括
  9. Sec
  10. ’an

Suppresses

  1. ELLOW
  2. ildo
  3. upported
  4. OnInit
  5. ыва
  6. ancellable
  7. zwarte
  8. isposable
  9. ellig
  10. <![