INDEX

Explanations

states, actions, and outcomes

The neuron responds to third-person singular present-tense verbs ending in “s” (e.g. occurs, includes, appears, results, returns).

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

引用元

-1.79

多くの

-1.51

跻

-1.50

Both

-1.48

 تجاه

-1.46

and

-1.45

具体的な

-1.42

籏

-1.42

There

-1.41

 рукой

-1.39

POSITIVE LOGITS

1.66

1.59

1.55

 today

1.52

’

1.36

 nettsted

1.35

外線

1.33

 samfun

1.32

 pazi

1.30

 аккура

1.24

Activations Density 0.096%