INDEX

Explanations

The neuron responds to long, multi-syllable technical or scientific terms (e.g. polysyllabic domain-specific words).

descriptor followed by technical noun

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

為に

-1.60



-1.36

あー

-1.34

嫱

-1.28

で使

-1.28

.«

-1.27

ほら

-1.27

鼩

-1.24

sandwich

-1.23

 noved

-1.23

POSITIVE LOGITS

ttää

1.22

<bos>

1.17

これが

1.13

This

1.12

 conferencias

1.11

 /\.(

1.07

rado

1.06

first

1.05

in

1.04

dieron

1.04

Activations Density 0.400%