INDEX

Explanations

actions, uses, and outcomes

The neuron activates on the main “action” verbs or predicate phrases that describe what an agent or system does (e.g. build, provide, employ, produce, perform).

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

voo

-1.26

艄

-1.21

錾

-1.15

申し訳

-1.13

ьере

-1.12

大人気

-1.12

 etek

-1.11

 ménages

-1.10

 HANS

-1.09

 mân

-1.09

POSITIVE LOGITS

 their

2.09

 themselves

1.68

his

1.32

кладки

1.30

It

1.30

 Perhaps

1.25

 Again

1.24

1.23

not

1.21

1.20

Activations Density 0.357%