"some" in "some kind" and "at some point"
A neuronOne of thousands of tiny dials inside each layer. It reads the text so far and produces a number, its activation, that nudges what the model writes next. in layerOne step in the stack the text passes through. Qwen3-0.6B has 28 layers, each with 3,072 neurons. 20 of Qwen3-0.6B. Its strongest snippets are very consistent: they mostly peak on the same token. It is featuredA neuron picked for the full treatment: steering outputs and a written explanation. About 3,000 of the 86,000 neurons are featured., so it has steering outputs.
What this neuron does
This neuron fires on the word "some" when it means an unspecified amount or kind, such as "some kind of", "at some point" or "in some way". Turned up, the model kept producing "some kind of some sort of".
Written by Claude from this neuron's activations and steering outputs. Confidence: high. Think it is wrong? Flag it (opens a prefilled issue on GitHub).
Fires most on
The text windows where this neuron's activationThe number a neuron produces for a token. Large positive means it fired hard; negative means it pushed the other way. was highest. Amber marks positive activation; the outlined tokenA piece of text the model reads at once: often a word, sometimes part of a word or a punctuation mark. is the peak. Hover a token for its strength.
-
protein production, they still involve proteins. So is it safe to say that while RNA performs several functions in the cell, all of them are related to proteins in some way? I might have answered
-
when you actually do well and are responsible in school.⏎Things you may find useful to stay focused:⏎- Computer with word processor or spreadsheet program⏎- Planner of some kind⏎- Goals sheet⏎- Cell phone or some kind of alarm reminder⏎
-
out of 100. Which may not sound like much.⏎But consider that 1 out of 15,000 kids gets cancer of some kind, and think about the attention and research and money devoted to understanding and treating
-
. Here are two important reasons why:⏎The odds of needing long-term care are high:⏎- Approximately 70% of people will need long-term care at some point during their lifetimes after reaching age 65.*⏎- Approximately 8
-
As<|endoftext|>How to Write a Research Paper. Whether you're in a history, literature, or science class, you'll probably have to write a research paper at some point. It may seem daunting when you're just starting out
-
Because of this, the sun will lose its brightness, becoming black for summers hereafter" (LXXXI.2). This may refer to an eclipse of some sort, or perhaps even a complete breakdown of our stratosphere, the skies
SteeringAdding a push along a neuron while the model writes, to see how its output changes. Turning it up or down.
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.
Here is something I have been thinking about lately: the more that a company can use a lot of its resources, the more it can afford to take a larger and larger amount of a single, specific and very specific, and therefore, the more it can afford a smaller and more specific,
Most negative on
Where the activation went furthest below zero. Blue marks negative activation. Many neurons mean something different in each direction.
-
AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author.While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file
-
c.⏎For technical questions regarding this item, or to correct its authors, title, abstract, bibliographic or download information, contact: (Michael McNulty)⏎If references are entirely missing, you can add them using this form.<|endoftext|>Farmers
-
egg.⏎If you are reading this message, Please click this link to reload this page.(Do not use your browser's "Refresh" button). Please email us if you're running the latest version of your browser and you still see this message.⏎
-
second session through Eventbrite as well, if you are interested in attending both: https://scisatadapt2.eventbrite.ca⏎Please contact us if you have any questions & we hope to see you there!<|endoftext|>Last week a
Connections
Show on mapWhich neuronsOne of thousands of tiny dials inside each layer. It reads the text so far and produces a number, its activation, that nudges what the model writes next. in earlier layersOne step in the stack the text passes through. Qwen3-0.6B has 28 layers, each with 3,072 neurons. feed this one, and which later ones it feeds. Amber links excite, blue links suppress; longer bars are stronger.
Fed by earlier layers
- L17:2579 suppresses
- L18:1920 suppresses
- L19:2274 excites
- L15:1828 suppresses
- L18:2820 excites
- L15:976 suppresses
- L13:1730 suppresses
- L14:1107 suppresses
- L13:1051 excites
- L19:289 excites
- L17:2219 suppresses
- L15:3059 suppresses 'and' or 'or' joining the last list items
- L16:730 excites
- L10:2898 suppresses
- L17:2396 suppresses
- L18:2960 excites
- L12:3054 suppresses
- L13:1990 weakly suppresses
- L11:679 weakly excites
- L14:2571 weakly suppresses
- L15:2322 weakly excites
- L7:1729 weakly excites
- L12:886 weakly excites
- L6:16 weakly excites
Feeds later layers
- L27:1124 strongly excites
- L22:926 strongly excites
- L23:2246 excites
- L21:2124 excites
- L22:2836 excites
- L24:823 excites
- L26:677 excites
- L23:2193 excites
- L21:1725 weakly excites
- L23:2536 weakly excites
- L25:871 weakly excites
- L25:2301 weakly excites
- L23:2339 weakly excites
- L27:1593 weakly suppresses
- L21:2585 weakly excites
- L21:2859 weakly excites
- L25:1588 weakly suppresses
- L21:2515 weakly excites
- L24:1078 weakly excites
- L26:434 weakly excites
- L21:913 weakly suppresses
- L21:547 weakly excites
- L27:1069 weakly suppresses
- L21:1705 weakly excites
Activation histogramA chart of how often a neuron takes each activation value.
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.
Logit lensReading which output tokens a neuron pushes toward or away from, straight off its output weights, before the later layers have their say.
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 layersOne step in the stack the text passes through. Qwen3-0.6B has 28 layers, each with 3,072 neurons. instead.
Promotes
- some
- some
- Some
- Some
- somehow
- 某种
- somewhere
- 一些
- somew
- SOME
Suppresses
- 率达
- ::__
- 度过
- amment
- 情况下
- unct
- 及以上
- двига
- .defaultProps
- .AnchorStyles