INDEX

Explanations

what i mean

This neuron detects self-referential clarification phrases such as “what I mean” or “what you mean.”

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

 étrang

-1.73

 autre

-1.66

 ralenti

-1.57

 refroid

-1.57

 duração

-1.48

 déchir

-1.47

 réve

-1.42

 échou

-1.42

 coû

-1.41

 précip

-1.41

POSITIVE LOGITS

and

2.02

 with

1.76

 although

1.69

 used

1.65

if

1.55

 which

1.45

set

1.30

)”.

1.27

 between

1.25

 uses

1.22

Activations Density 0.074%