INDEX

Explanations

shows concluding points

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

 面包

-1.00

🎇

-0.99

gefühl

-0.98

Ǟ

-0.96

Descriere

-0.94

 compagnon

-0.94

kommt

-0.94

pozorn

-0.93

 Finlande

-0.93

🌜

-0.91

POSITIVE LOGITS

 shows

4.25

 demonstrates

3.77

 indicates

3.67

 نشان

3.38

 Shows

3.22

shows

3.20

 demonstrate

3.17

 indicating

3.13

 illustrates

3.08

Shows

3.08

Activations Density 0.182%