INDEX

Explanations

phrases that indicate personal openness or disclosure

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

hiba

-0.08

ucas

-0.07

rosso

-0.07

ivol

-0.07

 Goldberg

-0.06

enk

-0.06

bai

-0.06

xca

-0.06

Äĥr

-0.06

bate

-0.06

POSITIVE LOGITS

 about

0.10

on

0.07

incinn

0.07

 tentang

0.07

how

0.07

 scoring

0.07

about

0.07

 Hers

0.07

oose

0.06

 ABOUT

0.06

Activations Density 0.003%