INDEX

Explanations

words starting with "th"

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

THROUGH

-1.28

Through

-1.26

 Through

-1.23

Via

-1.23

through

-1.16

Via

-1.12

Thru

-1.09

via

-1.07

 THROUGH

-1.07

thru

-1.05

POSITIVE LOGITS

th

0.82

 throw

0.73

 threw

0.72

 though

0.67

th

0.66

 thought

0.62

 throws

0.58

 throwing

0.58

 thrown

0.57

thu

0.56

Activations Density 0.106%