INDEX

Explanations

java code contexts

The neuron flags occurrences of static import statements in code.

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

and

-2.16

 Also

-1.91

 There

-1.68

but

-1.63

On

-1.58

</strong>

-1.56

It

-1.45

 begin

-1.43

not

-1.42

POSITIVE LOGITS

裢

2.19

纒

1.95

翃

1.91

 coreana

1.91

of

1.88

 cuit

1.81

푔

1.78

 temb

1.77

箆

1.77

 negativa

1.71

Activations Density 0.006%