INDEX

Explanations

few or couple of

The main thing this neuron does is detect indefinite small-quantity expressions such as “a couple,” “a few,” “few,” and similar phrases.

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 hầu

1.13

 nhiều

1.09

 cầu

1.06

рока

1.05

ي

0.98

很多

0.98

λό

0.98

 kèm

0.97

л

0.97

ই

0.96

POSITIVE LOGITS

 dozen

1.63

dozen

1.48

 hundred

1.23

 ribu

1.16

辏

1.16

ture

1.15

ah

1.07

 wenigen

1.03

 Minuten

1.02

Minutes

1.02

Activations Density 0.115%