INDEX

Explanations

abbreviations defined in parentheses

The neuron detects biomedical abbreviations or acronyms (typically all-caps initialisms) such as “CAD,” “DM,” “HCC,” “CV,” 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

↵

-1.64

-1.56

for

-1.54

 with

-1.52

-1.45

 also

-1.44

fa

-1.41

not

-1.38

-1.37

POSITIVE LOGITS

ᓚ

1.50

prende

1.49

trying

1.46



1.46

 unless

1.45

fatter

1.43

 comenzaron

1.41

jourd

1.41

֊

1.41

쨩

1.40

Activations Density 0.424%