INDEX

Explanations

references to the rate at which an algorithm or process converges

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

Datuak

-0.69

rej

-0.60

 coloc

-0.55

modelBuilder

-0.54

claimer

-0.53

 Stande

-0.53

躇

-0.53

cass

-0.52

uke

-0.52

зумі

-0.52

POSITIVE LOGITS

 convergence

3.20

Convergence

2.78

 Convergence

2.77

convergence

2.73

 converge

2.72

 converges

2.70

 converged

2.45

 converging

2.42

conver

2.31

 convergent

2.30

Activations Density 0.014%