'to' after 'need' in 'what you need to know'
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. 5 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 lights up on the word 'to' in phrases like 'everything you need to know' or 'what you need to get started'. It goes quiet, and slightly negative, on 'do' or 'with' in 'what does this have to do with'. Turning it up didn't produce a clear theme in the sample 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 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.
-
the complexity of the case and what needs to be achieved in the movement of the patient's teeth. A palatal expander
-
setting. This repair method⏎The above information equips you with everything you need to know about FCAW as a new welder. The techniques and advantages and
-
if done correctly, but there are a few things you need to know before starting. Keep reading to learn more about how long composting takes and what you need to get started!⏎How Long Does It Take To Break Compost Down?⏎This is
-
. They support teachers by providing a map of progression from Graduate through Proficient and Highly Accomplished to Lead teacher. They help teachers work out what they need to learn in order to become the kinds
-
right ratio to get perfect composting results. Compostable materials are divided into browns and greens and must be added carefully. This article discusses everything you need to know about composting ratio and how it affects the process.⏎Carbon to Nitrogen
-
What would need to happen? What would I need to have?”⏎For many people, the answer lies in having what someone else has
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: is it related to the internet? What is the connection between the internet and the previous mention of the time of the virus? What is the connection between the internet and the previous mention of the time of the virus? What is the connection between
Most negative on
Where the activation went furthest below zero. Blue marks negative activation. Many neurons mean something different in each direction.
-
to produce structures that are more efficient to run, cutting ongoing costs and minimizing the carbon footprint.⏎What does this have to do with BrightBuilt Barn?⏎Well, it turns out that the Barn was also designed
-
does. Methane gas (CH4) traps about 34 times more heat within the earth’s atmosphere.⏎So what does worm manure have to do with dog waste?⏎Well, dog waste is not just full of bacteria that causes disease
-
components of the Pie Model in the Ruler vs. Pie Concept.⏎Webster defines Truth as Reality. But reality to whom? (What does this have to do with consumer research? We will get to that later.) We have often heard the
-
Parkinson’s Australia Inc., 2014). Recognising these signs and symptoms can aid early diagnosis and slow disease progression.⏎What’s dopamine got to do with it?⏎Dopamine is often referenced as the chemical in our brain that
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
- L4:538 suppresses
- L1:1229 weakly excites
- L2:586 weakly suppresses
- L3:25 weakly suppresses
- L4:2787 weakly suppresses
- L4:207 weakly excites
- L3:1274 weakly suppresses
- L2:410 weakly suppresses
- L4:25 weakly excites
- L0:547 weakly excites
- L2:2267 weakly excites
- L4:2048 weakly suppresses
- L0:2822 weakly excites
- L0:526 weakly excites
- L4:2168 weakly excites
- L0:1096 weakly suppresses
- L3:717 weakly suppresses
- L4:37 weakly suppresses
- L4:1555 weakly excites
- L4:1119 weakly excites
- L4:2345 weakly suppresses
- L2:655 weakly suppresses
- L4:1052 weakly excites
- L2:2343 weakly excites
Feeds later layers
- L10:2075 suppresses
- L15:2494 suppresses
- L12:2737 suppresses
- L12:791 suppresses
- L24:1877 weakly suppresses
- L16:173 weakly suppresses
- L11:2286 weakly suppresses
- L8:1065 weakly suppresses
- L11:1173 weakly suppresses
- L15:2393 weakly excites
- L9:1737 weakly suppresses
- L6:2880 weakly suppresses
- L11:2253 weakly excites
- L6:201 weakly excites
- L18:684 weakly suppresses
- L12:1193 weakly suppresses
- L17:2403 weakly suppresses
- L7:1763 weakly suppresses
- L23:1625 weakly excites
- L8:1890 weakly suppresses
- L12:1637 weakly excites
- L14:871 weakly suppresses
- L12:1625 weakly excites
- L20:1937 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
- oui
- acro
- 最强
- unconditional
- unami
- -strong
- minimum
- strongest
- 坚强
- vars
Suppresses
- idden
- ạc
- 肤色
- 的文化
- 摛
- karış
- matrices
- ancest
- _matrices
- 的一面