INDEX

Explanations

code delimiters

The neuron fires on tokens that are parts of quoted string literals (especially capitalized names, labels, or identifiers).

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

are

-1.76

</h2>

-1.64

 zonder

-1.49

-1.38

 februari

-1.35

WithIdentifier

-1.33

 Hưng

-1.29

*</

-1.25

netje

-1.16

 Brief

-1.16

POSITIVE LOGITS

 only

1.76

很多的

1.52

 even

1.49

Only

1.45

 mumbled

1.43

ではまた

1.43

や

1.38

とある

1.34

 teilweise

1.33

といっても

1.33

Activations Density 0.123%