end-of-document boundary token

What this neuron does

This neuron responds most strongly at the invisible marker that signals the end of one web document and the start of another, especially when the text before it stops mid-thought. Turning it up gave wandering, unrelated text, so the steering does not show a clear effect.

Written by Claude from this neuron's activations and steering outputs. Confidence: medium. Think it is wrong? Flag it (opens an email to the maker with the neuron already filled in).

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

   an error has occurred. The ldexpf() and ldexpl() functions do not set errno.⏎See attributes(5) for descriptions of the following attributes:<|endoftext|>The

2. #2 peak

   phases to spot malfunctions and flaws. It often includes stress testing to estimate the product lifespan. The end goal is a product that meets the client’s quality specifications<|endoftext|>Diploid cells (WI-38, M

3. #3 peak

   enforcement agencies keep of known "disapproved" files that they want to search people's hard drives for. In these applications, keeping a database of the actual file<|endoftext|>Pre

4. #4 peak

   formulas and the questions. Students who work on their concepts would always excel in mathematical problems when compared to those who focus on memorizing solutions. Problems like limit solving<|endoftext|>How to dissolve polygons spatially in GRASS GIS⏎You can use the v

5. #5 peak

   , the range and severity of symptoms and physical characteristics may vary from case to case.⏎LEOPARD is an acronym for the characteristic abnormalities associated with the disorder: L stands for (L)entigines (multiple black or dark brown spots on

6. #6 peak

   mann machines (RBM), deep belief networks (DBN) and deep convolutional neural networks (CNN). In recent years, CNN-based methods have gained popularity<|endoftext|>typedef unsigned int size_t;⏎This type is used to hold size definitions. Since

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

   I walk to the beach.⏎- Action 2⏎We collect all sorts of seashells.⏎- Action 3⏎I spot something dark and shiny.⏎- Action 4⏎I grab it before a wave splashes it away.⏎-

2. #2 peak

   5 to 88 percent. MENA achieved the same increase in 2 decades: from 1990 to 2010.2⏎However, it is important to note that regional statistics tend to mask the fact that

3. #3 peak

   those declarations. Humans read code, and humans infer things from the code that will be necessary to work on it.⏎typedef void * model_3d_ref;⏎model_3d_ref load_model_from_disk(const char * absolute_file_path);⏎void

4. #4 peak

   modification, would have been destructive of the theory and would have made the<|endoftext|>It is not good for the health of a tree to build up soil around its base.⏎In the Chicago suburbs and near Galena, I have planted trees in soil that

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.