INDEX

Explanations

definite article followed by a noun

The neuron strongly activates on numeric measurements expressed with decimal points (e.g., “624.9432,” “500.4466,” etc.).

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 Anschließend

0.77

Sp

0.74

PAD

0.74

NELL

0.73

◎

0.72

･

0.71

하지만

0.69

WI

0.69

Wr

0.68

GALL

0.68

POSITIVE LOGITS

 politik

0.78

 eyes

0.76

 injustices

0.75

 bore

0.72

 injustice

0.70

 episodes

0.69

boc

0.68

 behaviour

0.68

 ansatz

0.68

 comics

0.67

Activations Density 0.000%