INDEX

Explanations

introducing explanatory examples

This neuron detects phrasing that introduces an illustrative example (e.g. “take a simple example,” “consider the following example,” “here is an example,” etc.).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

ventions

-1.50

to

-1.44

 also

-1.36

 gainera

-1.34



-1.30

giveness

-1.27

atical

-1.27

 improvements

-1.27

 include

-1.25

≤

-1.23

POSITIVE LOGITS

 where

1.43

1.38

of

1.34

 indisponible

1.32

 玄関

1.32

 cuyos

1.30

 Ако

1.25



1.25

τέλε

1.25

 Тогда

1.24

Activations Density 0.160%