INDEX

Explanations

syncing data

The main thing this neuron does is detect the word “sync” (and its inflected or derived forms like “synchronize,” “synchronizing,” “synchronized,” etc.).

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

-1.88

 или

-1.72

 hóspedes

-1.70



-1.58

::*;

-1.57

-1.55

 enemigos

-1.55

舠

-1.48

POSITIVE LOGITS

以上です

1.91

※※

1.90

＊＊＊＊

1.90

ではでは

1.84

 between

1.82



1.77

個人的

1.77

＝＝＝＝

1.72

➢

1.67

－－－－

1.66

Activations Density 0.009%