INDEX

Explanations

a/an + adjective

The neuron activates on numeric tokens (e.g. numbers, especially multi-digit or decimal figures).

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

 оригі

-1.66

 should

-1.48

 faltan

-1.41

原來

-1.41

 with

-1.31

-1.30

 дії

-1.30

誒

-1.28

 좋은

-1.28

البته

-1.27

POSITIVE LOGITS

⌚

1.25

 entangled

1.25

 emphasis

1.21

 αλλά

1.21

 poet

1.21

☑️

1.20

.,

1.20

</strong>

1.20

?,

1.20

Activations Density 0.057%