INDEX

Explanations

contractions and possessives

The neuron activates on the fragments of common English contractions (e.g. “isn” in isn’t, “‘m” in I’m, “‘ve” in I’ve).

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

—

-1.58

to

-1.53

 with

-1.16

The

-1.16

Према

-1.13

or

-1.08

—

-1.07

значно

-1.05

-1.02

POSITIVE LOGITS

 paño

1.35

 leopardo

1.29



1.25

archiwizowane

1.23

 abandonné

1.20

 miniaturka

1.20

 kanal

1.19

 barras

1.17

 видели

1.17

 adesivo

1.16

Activations Density 0.104%