INDEX

Explanations

Go type and function declarations

The neuron fires on numeric literals (numbers) in the code.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 belki

-1.98

 verhind

-1.96

礻

-1.93

 cierta

-1.84

真好

-1.77

рабаты

-1.75

 úpl

-1.72

 pomaga

-1.71

ouldering

-1.68

真的很

-1.67

POSITIVE LOGITS

<bos>

2.30

Value

1.45

 lama

1.41

zsef

1.40

kệ

1.39

 rita

1.38

したのは

1.36

one

1.36

 tanya

1.35

set

1.34

Activations Density 0.024%