INDEX

Explanations

for example or etc

The neuron fires strongly on technical or scientific nouns—especially those naming components or sources in environmental/mixture contexts (e.g. “sources,” “stream,” “molecules,” “tree”).

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

mainly

-1.44

basically

-1.40

XIX

-1.25

apparently

-1.22

primarily

-1.21

mostly

-1.20

»:

-1.19

essentially

-1.15

 ἐν

-1.13

Mainly

-1.13

POSITIVE LOGITS

or

3.53

etc

2.33

 например

1.65

 یا

1.57

或者

1.52

 или

1.49

 might

1.36

etc

1.36

或

1.35

或是

1.34

Activations Density 0.113%