INDEX

Explanations

Here's a breakdown of why

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

4.50

‑

3.09

-...

2.65

2.57

-<

2.48

2.44

‐

2.41

-'

2.40

-*

2.38

-{

2.35

POSITIVE LOGITS

༈

1.25

 PARTICULAR

1.15

힣

1.11

 PRINC

1.08

kbar

1.08

 INTERNAL

1.07

 ADDITIONAL

1.07

闿

1.05

 mapDispatch

1.05

पेयी

1.05

Activations Density 0.241%