chinese 的 particle before a noun
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 mixed: they peak on different tokens. 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 mostly fires on the Chinese word 的 (and similar linking words) when it comes before a noun, especially in Chinese academic text. It is much weaker on English, where it fires on a few words like 'insidious'. Turning it up or down did not give a clear change in the model's English output.
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.
-
法、分析法来进行研究,探讨泰国学生在学习汉语助动词中的困难及偏误情况,分析他们产生偏误的原因。我们发现经过
-
Chinese Auxiliary Verbs for Thai Learners Based on the Comparative Analysis⏎- 人文学院-已发表论文 ⏎“助动词”在汉语中具有较大的作用,使用率高,但是泰国学生经常在学习和使用汉语助
-
to describe a global TCP/IP network. So, you could also argue the Internet (<|endoftext|>このトピックには0件の返信が含まれ、1人の参加者がいます。2020年7月6日 at 9
-
, and procrastinating until the last moment can feel overwhelming to both students and parents. Additionally, it
-
��性病的⏎anachronism /əˈnækrəˌnɪz
-
If you've never experienced the monster that is depression, there is really no way for you to grasp its horrors. One of the things that makes depression so insidious is that it distorts one's self-perception and causes the belief that the
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: I want to be more involved in the world, and I want to be more involved in the world, and I want to be more involved in the world, and I want to be more involved in the world, and I want to be
Most negative on
Where the activation went furthest below zero. Blue marks negative activation. Many neurons mean something different in each direction.
-
and paying back on time that you can build up a good score.”⏎Happily, while In addition, three in ten people are worried<|endoftext|>Formation Of Plural Nouns⏎An English noun exhibits a two-way distinction of number: a singular
-
ten objects from museum collections in Greater Manchester⏎A compass used in a PoW escape, a railway station sundial and part of the world's first commercial computer: objects that tell a history of
-
ist (pointillist) techniques.⏎He also studied with Henri de Toulouse-Lautrec, and he exhibited works at Cafe Tambourin with Louis Anquetin, Emile Bernard, and Toulouse-Laut
-
they can get advantages anytime.⏎Why Expertsmind for assignment help⏎- Higher degree holder and experienced experts network⏎- Punctuality and responsibility of work⏎- Quality solution with 100% plagiarism
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
- L4:300 strongly excites
- L2:1473 excites
- L12:1058 excites
- L16:1433 excites
- L3:1774 excites
- L6:1063 excites
- L17:3043 excites
- L7:111 excites
- L18:1265 excites
- L6:1233 excites
- L26:1776 excites
- L15:286 excites
- L21:2486 excites
- L9:366 excites
- L9:240 excites
- L8:2755 excites
- L4:2743 excites
- L5:1921 excites
- L8:252 suppresses 'of' after 'each' or 'one' or 'all'
- L18:602 excites
- L13:1196 excites opening paren before an acronym
- L14:118 suppresses
- L14:280 weakly suppresses
- L7:210 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
- ⏎⏎
- urga
- `,
- uddenly
- 郎
- loạt
- ⏎ ⏎
- `,⏎
- бег
- URITY
Suppresses
- _Reference
- enschaft
- 的真实性
- ance
- 性的
- raising
- 一楼
- aticon
- unas
- 人民服务