INDEX

Explanations

speaking with emphasis

The neuron fires strongly on words signaling loud vocalizations or calls (e.g. “loud,” “voice,” “shouting,” “cry,” “called”).

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

 (!__

-1.50

(!__

-1.49

we

-1.45

 montagnes

-1.45

providedIn

-1.43

൮

-1.41

 pueblos

-1.38

ఝ

-1.36

voer

-1.35

特斯拉

-1.33

POSITIVE LOGITS

</i>

1.69

 Whereas

1.63

1.52

 Compared

1.47

 While

1.47

1.43

 Traditionally

1.41

 Historically

1.39

而

1.31

 Unlike

1.30

Activations Density 0.002%