INDEX

Explanations

interval notation [0,1]

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

</b>

-1.52

 beijo

-1.47

%);

-1.41

 dispozici

-1.41

))->

-1.39

͊

-1.35

𝒐

-1.34

cannot

-1.33

 undangan

-1.31

Портали

-1.30

POSITIVE LOGITS

1.67

1.61

but

1.60

</h3>

1.49

1.45

鐽

1.34

1.28

1.24

↵↵↵

1.23

 relics

1.20

Activations Density 0.020%