INDEX

Explanations

home ofshadow of

The neuron selectively responds to tokens that are part of proper names, acronyms, or other specialized technical identifiers (e.g. “home of,” “erasable,” “CD,” “U(1),” “EMG,” “cancel,” “SetCompatibleTextRenderingDefault”).

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

我が家

-1.02

 حتی

-1.00

״

-1.00

 aktiviert

-0.98

 even

-0.98

 рішення

-0.97

lė

-0.97

 Portail

-0.96

WARDS

-0.96

 він

-0.96

POSITIVE LOGITS

 which

1.02

したいと思います

1.02

了一遍

0.96

させていただきました

0.96

してました

0.95

 modelleri

0.93

 when

0.93

聆

0.92

しましたが

0.91

صبحت

0.91

Activations Density 0.133%