INDEX

Explanations

introduction

The neuron flags tokens in publication‐metadata or disclaimer statements (e.g. disclosures, errata, conflict‐of‐interest notices, etc.).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 Nedir

-1.27

 normaux

-1.16

 montrent

-1.09

 estimés

-1.09

 Ahora

-1.08

 squadre

-1.07

 définiti

-1.06

*}[

-1.06

 från

-1.05

 arrivent

-1.05

POSITIVE LOGITS

and

2.14

 with

1.20

for

1.09

انتهای

1.09

ɲ

1.04

剷

1.03

tradu

0.98

 Gläub

0.96

 bewegt

0.95

 вули

0.94

Activations Density 0.002%