INDEX

Explanations

events happening before other events

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

 wakes

-0.08

iez

-0.08

fr

-0.07

anders

-0.07

oi

-0.07

etro

-0.07

Est

-0.07

ogh

-0.07

appings

-0.07

rs

-0.07

POSITIVE LOGITS

 decency

0.13

 anyone

0.12

 anybody

0.11

rita

0.11

 succ

0.10

Maj

0.10

aba

0.10

can

0.10

 consciously

0.10

WARN

0.10

Activations Density 0.003%