defining or explaining concepts

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

1.08

the

1.01

two

0.98

0.96

 high

0.90

 local

0.88

des

0.82

al

0.81

POSITIVE LOGITS

DanhMucSP

1.96

<unused1207>

1.93

PAssignment

1.91

<unused30>

1.90

kepsilon

1.90

<unused790>

1.90

1.89

<unused513>

1.89

𒈩

1.88

<unused1183>

1.87

Activations Density 0.992%