INDEX

Explanations

resigned or resignation

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

KA

-0.09

holm

-0.09

Recovered

-0.09

ield

-0.09

 Emin

-0.09

isp

-0.09

 Assigned

-0.08

adera

-0.08

 breakthrough

-0.08

:type

-0.08

POSITIVE LOGITS

 resign

0.37

 resignation

0.35

 quitting

0.29

 resigned

0.27

 quit

0.25

 quits

0.24

 Quit

0.22

 departing

0.19

 Leaving

0.19

 leaving

0.18

Activations Density 0.361%