INDEX

Explanations

question/answer starters

This neuron responds most strongly to high-frequency function words (e.g. “a,” “be,” “can,” “which,” “would,” etc.).

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

 then

-1.30

 this

-1.30

 their

-1.20

 three

-1.13

 other

-1.12

 only

-1.08

Sábado

-1.06

 five

-1.02

 both

-1.02

 using

-1.01

POSITIVE LOGITS

如果你

1.20

пель

1.16

м

1.13

)});

1.12

凭借

1.11

тан

1.11

If

1.11

ʏ

1.09

 ſer

1.08

要想

1.07

Activations Density 0.000%