INDEX

Explanations

to + verb

The neuron responds to the verb “look” when it’s used in constructions like “look(s) to” indicating intent or direction.

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

dersfield

-1.93

為了

-1.91

鱻

-1.88

ischof

-1.81

還會

-1.77

一開始

-1.77

應該是

-1.75

pkins

-1.75

 wszystkie

-1.75

POSITIVE LOGITS

an

2.52

1.84

al

1.81

 oder

1.63

at

1.63

ان

1.62

 fella

1.61

re

1.56

",

1.55

Activations Density 0.010%