INDEX

Explanations

at or in followed by location or time

The neuron detects prepositions (like “at” or “in”) that introduce a location or venue.

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

쨩

-1.45

actéristi

-1.43

 cuchara

-1.41

 vone

-1.38

 durante

-1.38

髫

-1.33

 !!!!

-1.30

 DURING

-1.30

 medición

-1.28

ḿ

-1.28

POSITIVE LOGITS

 with

2.23

of

2.11

at

1.91

on

1.78

 will

1.77

 before

1.69

in

1.64

 after

1.53

 four

1.46

Two

1.41

Activations Density 0.001%