ordinal suffix "nd" after numbers
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 reacts to the "nd" in dates and ordinal numbers like "22nd" or "2nd". It seems to dislike the digit "0" in copyright years like 2012. Turning it up did not clearly change the writing.
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.
-
not have to wait until the 22nd century for a usable equivalent of Star Trek’s universal translator, and we can also
-
and captured by the U.S. Army 2nd Division in Korea, during the Korean War. A convoy of jeeps and trucks
-
air purifiers. Learn more about tech gadgets that will protect your home against air pollutants<|endoftext|>Active and Passive Voice – Tips for Content Writing⏎February 22nd, 2012 by Donna Mathews⏎The use of active voice
-
Category Archives: Maths 1st ESO⏎Download the course⏎Download the problems and do at least four of them: (Deadline March, 22nd) 1st and 2nd ESO 3rd and 4
-
: How to teach writing like you mean it (2nd ed.). Newark, DE: International Reading Association.⏎- Writing activities that help students
-
key signature of A major and makes changes to a lot of 2nd fingers.⏎A new scale: A major⏎Add it to your 'warm-up
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 have been thinking about lately. I have been thinking about lately. I have been thinking about lately. I have been thinking about lately. I have been thinking about lately. I have been thinking about lately. I have been thinking about lately
Most negative on
Where the activation went furthest below zero. Blue marks negative activation. Many neurons mean something different in each direction.
-
ation | Video⏎- Mars Rover Curiosity: Mars Science Lab Coverage⏎- 1st Photos of Mars by Curiosity Rover (Gallery)⏎Copyright 2012 SPACE.com, a TechMediaNetwork company. All rights reserved. This
-
Delivery: Yes⏎- Store Delivery: Yes⏎Prices are for internet purchases only. Prices and availability in WHSmith Stores may vary significantly⏎© Copyright 2013 - 2017 WHSmith and its suppliers.⏎WHSmith
-
.⏎from Wiktionary, Creative Commons Attribution/Share-Alike License⏎- n. An underwater explorer.⏎from WordNet 3.0 Copyright 2006 by Princeton University. All rights reserved.⏎- n. an underwater swim
-
parts of Scandinavia and Finland with the Kola Peninsula in Russia, traditionally inhabited by the Sami people.⏎from WordNet 3.0 Copyright 2006 by Princeton University
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
- L4:1648 suppresses
- L1:35 suppresses
- L21:956 excites
- L5:3052 weakly excites weighing, counting and measuring
- L16:1818 weakly excites "two" and the link between two things
- L2:1436 weakly excites
- L20:2765 weakly excites
- L6:1346 weakly suppresses
- L2:2900 weakly excites
- L1:2787 weakly excites
- L27:422 weakly excites
- L13:71 weakly suppresses
- L18:3068 weakly excites
- L12:2869 weakly suppresses
- L3:279 weakly suppresses
- L3:230 weakly excites
- L21:2731 weakly excites
- L2:2912 weakly excites
- L23:1708 weakly excites
- L3:2064 weakly suppresses
- L15:2022 weakly excites
- L27:2197 weakly excites
- L11:2528 weakly excites
- L6:1465 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
- ście
- ega
- AXB
- ref
- زي
- eger
- ebx
- ILL
- 上级
- bull
Suppresses
- '</
- BeginInit
- 把他
- ipated
- }\"
- //}}
- 扫一
- ''.
- 岁时
- 饹