INDEX

Explanations

activity prediction

The neuron activates on numeric tokens (i.e. any standalone numbers in the text).

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

getItem

-0.81

 continues

-0.79

ervi

-0.78

 Fischer

-0.76

"))

-0.73

ändler

-0.73

 Hartman

-0.71

lanterna

-0.71

ive

-0.71

Easter

-0.71

POSITIVE LOGITS

cies

0.84

 quieras

0.83

ONS

0.79

𝒌

0.77

zija

0.75

atics

0.75

rom

0.74

unner

0.73

 cheia

0.73

 rộng

0.72

Activations Density 0.014%