INDEX

Explanations

naming and simplifying concepts

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

 remark

-0.12

 comments

-0.12

 remarks

-0.12

 comment

-0.12

 sentiments

-0.12

 sentiment

-0.12

 replies

-0.11

 Response

-0.11

 Comments

-0.11

 disagreement

-0.11

POSITIVE LOGITS

 straightforward

0.15

 succinct

0.11

 simplest

0.10

 naming

0.10

 simpl

0.10

Shortcut

0.10

 trium

0.09

 shocking

0.09

 shorthand

0.09

 simplified

0.09

Activations Density 0.042%