INDEX

Explanations

strange colours, strange problem

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

tridge

-0.11

ç§

-0.10

ture

-0.10

CAPE

-0.09

 Mitchell

-0.09

 Scatter

-0.09

rum

-0.09

illa

-0.09

isclosed

-0.09

 promin

-0.09

POSITIVE LOGITS

 myster

0.15

 strange

0.13

 inexp

0.13

 Strange

0.12

 strangely

0.12

Strange

0.11

 weird

0.11

 suddenly

0.11

 vanished

0.10

 oddly

0.10

Activations Density 0.080%