INDEX

Explanations

reason why

The neuron detects words and phrases associated with quantitative reporting or measurement contexts (e.g. counts, prevalence, reported, risk factors, objectives) that introduce or accompany numerical data.

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

ValueGeneration

-1.30

to

-1.25

initComponents

-1.23

beca

-1.23

 том

-1.21

 когда

-1.21

paramInt

-1.20

>");

-1.19

̴

-1.17

__*/

-1.13

POSITIVE LOGITS

 gewerb

1.27

 îles

1.17

either

1.16

多かった

1.13

 Chinois

1.12

 取り付け

1.08

at

1.07

 either

1.04

batal

1.03

 Eigentü

1.03

Activations Density 0.252%