INDEX

Explanations

sitting in a location

The neuron fires on words that name or mark physical locations or spatial settings (e.g. “stage,” “room,” “board,” “facility,” etc.).

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

↵↵↵↵↵↵↵↵↵↵

-1.44

↵↵↵↵↵↵↵↵↵↵↵↵

-1.41

↵↵↵↵↵↵↵↵↵↵↵↵↵

-1.38

 aastal

-1.38

irr

-1.35

↵↵↵↵↵↵↵↵

-1.34

?}

-1.30

ciun

-1.24

↵↵↵↵↵↵↵↵↵↵↵

-1.23

↵↵↵↵↵↵↵

-1.23

POSITIVE LOGITS

of

1.41

 with

1.35

for

1.34

one

1.29

who

1.27

struttura

1.27

 from

1.14

at

1.05

 refused

1.03

 positioned

1.02

Activations Density 0.050%