INDEX

Explanations

Starts of specific phrases

The neuron responds strongly to punctuation delimiter tokens (e.g. commas and underscores).

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

 respective

-1.52

很赞

-1.49

istocene

-1.47

 événement

-1.44

可愛

-1.40

 zprávy

-1.38

現在は

-1.34

そして

-1.34

 тільки

-1.34

 pregunto

-1.32

POSITIVE LOGITS

 that

2.23

<bos>

1.89

that

1.63

 which

1.59

the

1.55

 That

1.49

 estekak

1.36

it

1.30

寓

1.29

we

1.28

Activations Density 0.000%