INDEX

Explanations

The main thing this neuron does is detect numeric measurements and values (e.g. dimensions, quantities, and unit‐based figures).

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 متنوع

0.91

 अनेक

0.89

 विश्वविद्यालयों

0.84

多种

0.83

 समितियों

0.83

 universidades

0.80

ospitals

0.80

 परिवारों

0.79

 المختلفة

0.79

ﻋ

0.78

POSITIVE LOGITS

the

1.93

the

1.43

it

1.38

 said

1.30

該

1.26

fhe

1.23

该

1.22

teh

1.18

ihe

1.10

one

1.05

Activations Density 1.110%