INDEX

Explanations

programming constructs and types

The neuron fires on type‐name tokens used as generic parameters (i.e. CamelCase identifiers inside “<…>”).

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

to

-2.05

on

-1.73

-1.44

for

-1.41

is

-1.36

in

-1.34

at

-1.33

 your

-1.31

 with

-1.30

-1.23

POSITIVE LOGITS

 lorsqu

1.44

 ومل

1.39

фотография

1.38

觴

1.38

Locals

1.38

lemény

1.35

 extinción

1.35

1.34

 május

1.33

辍

1.33

Activations Density 0.023%