INDEX

Explanations

Questions

The neuron fires on the literal token “Type,” i.e. when the document labels or annotates a property or attribute as “Type.”

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

NÍ

-1.21

pernic

-1.17

RRRR

-1.13

rdoba

-1.13

ksiyon

-1.11

ERON

-1.09

кономі

-1.08

ಯ

-1.06

 redaktion

-1.06

DREN

-1.05

POSITIVE LOGITS

 there

1.26

1.22

などと

1.21

</strong>

1.13

ポイントは

1.08

1.05

1.04

理由は

1.02

not

1.01

</

1.01

Activations Density 0.000%