INDEX

Explanations

noun followed by preposition

The neuron strongly activates on longer, multi-syllable content words (e.g. “released,” “longevity,” “performance”), effectively detecting unusually long words.

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

 inconce

-1.26

 abomin

-1.23



-1.23

⮑

-1.15

 commentaire

-1.15

 abusi

-1.14

 kazak

-1.13

 infinites

-1.12

냑

-1.11

་

-1.09

POSITIVE LOGITS

of

1.41

1.11

on

1.10

in

1.10

if

1.05

for

1.02

 from

1.02

 those

1.02

new

1.01

the

1.01

Activations Density 0.099%