INDEX

Explanations

objects

The neuron specifically detects occurrences of the word “objects.”

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

."

-1.86

”？

-1.75

nisses

-1.73

 evidenced

-1.66

rzej

-1.58

ONDER

-1.56

娌

-1.55

 nghiệ

-1.53

as

-1.52

.'”

-1.47

POSITIVE LOGITS

 cotidiano

1.74

 antigas

1.70

 that

1.68

当年

1.63

 として

1.57

いいのか

1.55

フェス

1.49

 these

1.48

鵡

1.48

几分

1.45

Activations Density 0.017%