full stops in running text
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. 27 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 responds to full stops and dashes in ordinary prose, mostly where a sentence or clause ends. Its signals are much weaker than the other neurons here. Turning it down or up produced repeated filler words, so the steering does not show what it does.
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 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.
-
Trial to Unlock the Full Review⏎Why Lesson Planet?⏎Find quality lesson planning resources, fast!⏎Share & remix collections to collaborate.⏎Organize your curriculum with collections. Easy!⏎Have time to be more creative & energetic with your students!⏎Students are
-
Stop breadboarding and soldering – start making immediately! Adafruit’s Circuit Playground is jam-packed with LEDs, sensors, buttons, alligator clip pads and more. Build projects with Circuit Playground in a few minutes with the drag-and-drop MakeCode
-
have the proper application to view it first. In case of further problems read the IDEAS help page. Note that these files are not on the IDEAS site. Please be patient as the files may be large.⏎As the access to this document
-
zipping the .tgz files using the terminal. Let’s get into these.⏎Note: For this particular tutorial, we will use the Ubuntu 22.04. However, the commands used here are same for all the distributions.⏎
-
Trial to Unlock the Full Review⏎Why Lesson Planet?⏎Find quality lesson planning resources, fast!⏎Share & remix collections to collaborate.⏎Organize your curriculum with collections. Easy!⏎Have time to be more creative & energetic with your students!⏎Students explore
-
completely new experimental access to quantum physics has been made at University of Konstanz. The team of scientists headed by...⏎Yersiniae cause severe intestinal infections. Studies using Yersinia pseudotuberculosis as a model organism aim to elucid
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: in the in the in in in in in in in in in in in in in in in in in in in in in in in in in in in in in in in in in in in in in in in in in in in in
Most negative on
Where the activation went furthest below zero. Blue marks negative activation. Many neurons mean something different in each direction.
-
דמה DooMaH (Ezekiel27:32)⏎and דממה DiMaMaH (IKings19:12)⏎all mean silence. דמום DahMOOM is silent, דמם
-
under CC BY-NC-SA⏎Given an integer , the exponential sum version of the Gauss sum modulo is defined by⏎and the quadratic residue version is defined by⏎where is the Jacobi symbol, which has values
-
this problem?⏎offers the opportunity for a statistical investigation⏎of confidence intervals and expectations at various levels of⏎There are three possible levels at which this problem can be⏎tackled, and how it is used depends on the levels of the students
-
rendered that Aaron was DUMBFOUNDED. DIM⏎perception is seen in words like דמדום DiMDOOM⏎("dim light" and "blackout"-- as in unconscious). טמום
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
- L19:68 strongly excites full stops closing a claim
- L23:509 strongly suppresses
- L13:44 strongly suppresses inside web addresses and citations
- L22:3045 strongly suppresses
- L18:178 excites
- L25:30 suppresses
- L15:184 excites
- L26:27 excites
- L22:14 excites
- L15:390 excites
- L20:59 suppresses
- L11:223 excites
- L22:67 suppresses
- L23:10 excites
- L19:1240 suppresses
- L10:669 suppresses
- L19:214 excites
- L21:201 excites
- L15:59 suppresses
- L19:1174 excites
- L24:353 excites decimal point inside a number
- L19:124 suppresses
- L14:140 suppresses
- L10:155 excites
Feeds later layers
None: this is the last layer.
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
- egrity
- stakes
- missions
- 不知不
- assadors
- plies
- 鲑
- hone
- reece
- DOCTYPE
Suppresses
- in
- for
- then
- that
- to
- according
- if
- otherwise
- the
- during