'this' after sentence-initial 'In'
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. 0 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 'this' in the opening phrase 'In this guide' or 'In this article', which is a common way for web articles to start. It is one of the strongest and most consistent neurons in this set. Turning it up did not produce any clear theme in the model's replies.
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.
-
you.⏎In this guide, I’ll go through all the main elements pertaining to SSL certificates to provide
-
of interest to you.⏎In this guide, we’ll go over all of this in detail. We’ll explore the
-
of foreign policy.”⏎In this “age of great migration,” more societies are becoming like the United States, with
-
’s heritage.⏎In this article, I’m going to look at the history of American culture and how that
-
Behind the Northern Lights⏎In this interactive session, Norwegian solar physicist Pål Brekke utilizes photos and video footage from
-
versus broadcasting nutrients is likely to become a much bigger conversation as Ontario works to reduce the impact farm phosphorous on the Great Lakes.⏎In this Corn School episode, Purdue University professor Tony Vyn offers insights on why he feels
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: how to make a circle of the day. I know that the circle is made by connecting the points. So, how to connect the points to make a circle. The points are the points of the day. The circle is the circle of
Most negative on
Where the activation went furthest below zero. Blue marks negative activation. Many neurons mean something different in each direction.
-
gardenia.⏎Insecticides kill off the whitefly’s natural predators. These predators include other insects and
-
the continued loss of habitat in parks across the US, this is worth a read. The full article can be accessed here.⏎Tags: New York Times, Insect Population Decline<|endoftext|>First posted on Myth & Moor in 2015
-
-BEVS is a promising alternative strategy to produce the functional and thermostable reverse transcriptase.⏎All Science Journal Classification (ASJC) codes⏎- Insect Science<|endoftext|>An account of the religion of
-
fr.wikipedia.org⏎Note: you should have a Insecta.pro account to upload new topics and comments. Please, create an account or
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
None: this is the first layer.
Feeds later layers
- L7:1220 excites
- L8:51 weakly excites
- L11:2894 weakly suppresses
- L15:194 weakly excites
- L12:2422 weakly excites
- L22:1127 weakly excites machine learning explanations, mid-sentence
- L2:2992 weakly suppresses
- L7:348 weakly excites
- L12:2316 weakly excites
- L11:1173 weakly excites
- L8:2299 weakly excites
- L12:1541 weakly suppresses
- L24:2376 weakly excites
- L3:2345 weakly suppresses
- L16:2787 weakly suppresses
- L9:1785 weakly excites
- L9:1210 weakly excites
- L13:2495 weakly suppresses
- L3:807 weakly suppresses comma inside dates and years
- L19:2466 weakly excites
- L24:2234 weakly suppresses
- L24:1901 weakly excites
- L6:1115 weakly suppresses
- L16:2073 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
- vido
- fer
- OMP
- Kimber
- 方式进行
- なもの
- ose
- 止
- away
- oir
Suppresses
- erializer
- icolor
- CharCode
- Endian
- iết
- >)
- Washer
- )")⏎⏎
- bitte
- bitten