the word 'about'

What this neuron does

This neuron lights up on the word 'about' in many senses: 'that's about all', 'about to try', 'about 99%', 'about as sexy as'. Turning it up made the model's text drift toward news-like and general wording, though the steered outputs are not very clearly on one theme.

Written by Claude from this neuron's activations and steering outputs. Confidence: high. Think it is wrong? Flag it (opens a prefilled issue on GitHub).

Fires most on

The text windows where this neuron's activation was highest. Amber marks positive activation; the outlined token is the peak. Hover a token for its strength.

1. #1 peak

   and francophone countries.⏎- Colonial rule⏎ASJC Scopus subject areas⏎- Economics and Econometrics⏎- Organizational Behavior and Human Resource Management<|endoftext|>- About ASTI⏎- Pay and Conditions⏎- ASTI Membership⏎- Operation of

2. #2 peak

   you can attach a label to it that says this home meets the HUD Code. Obviously there are several layers of bureaucracy involved in all of this, but that’s about all there is to it.⏎There are fines for failures to comply with the Code

3. #3 peak

   the family from then on.⏎That’s about all I’d say in a first go-round; go into other details (like

4. #4 peak

   transmission of electrical signals and also deliver energy to neurons and remove neuronal waste products. While appearing to be a little more interesting than originally thought, glia still seemed about as sexy as wire insulation, food delivery and

5. #5 peak

   the Museum of Connecticut Glass hosted "It's About Connecticut Glass," an exchange between local historical societies, representatives from Connecticut glass works,

6. #6 peak

   Overall, these tests have excellent detection rates (~99%) with very low false-positive rates (~0.2%). In other words, these tests are about 99.0% sensitive and 99.8% specific.⏎

Steering

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.

Most negative on

Where the activation went furthest below zero. Blue marks negative activation. Many neurons mean something different in each direction.

1. #1 peak

   Bethel and Dan, whose golden bull-calf statues they declared idolatrous.⏎In reaction<|endoftext|>Like Father, Unlike Son⏎The Battle of Tours in 732 was a turning point in the wars against the Umayyad Cal

2. #2 peak

   , January 2003, December 2006⏎Part I: The Umayyad Caliphate (661 – 750 CE)⏎5 Tibet on the Eve of the Arrival of the First Muslim

3. #3 peak

   much more!⏎Discover the world of gorillas and other primates their remarkable intelligence and social behavior⏎66 pages; ISBN 9780756668082⏎, or download in⏎More Ju

4. #4 peak

   that usually is followed by an implication that could be generalized for a broader extrapolated definition. This particular one is derived from the good old Spring and Autumn period (772BC to 403BC) and the Warring States period

Which neurons in earlier layers feed this one, and which later ones it feeds. Amber links excite, blue links suppress; longer bars are stronger.

Activation histogram

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 lens

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 layers instead.