INDEX

Explanations

interpret information confirming beliefs

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

for

1.12

Г

0.98

voor

0.95

â

0.91

Â

0.90

CN

0.90

Ã

0.90

̀

0.90

ونک

0.88

MO

0.88

POSITIVE LOGITS

 isomorphisms

1.56

 homomorphism

1.46

 isomorphism

1.45

≐

1.45

 hypotheses

1.45

논

1.41

ląd

1.38

 orthogonality

1.36

ાડ

1.35

 axioms

1.34

Activations Density 0.010%