INDEX

Explanations

names

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

æļ¾

-0.31

,GL

-0.28

geries

-0.27

plx

-0.27

ronic

-0.26

plen

-0.26

ryo

-0.26

plet

-0.25

ACH

-0.24

hold

-0.24

POSITIVE LOGITS

 Camel

0.42

èĤ¯

0.39

 kenn

0.35

Kenn

0.33

Di

0.33

 Kennedy

0.31

Di

0.31

JFK

0.30

acier

0.29

Ly

0.28

Activations Density 0.964%