INDEX

Explanations

describing qualities

The neuron detects descriptive adjectives (i.e. words that modify nouns by attributing qualities).

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

 about

-1.58

かは

-1.43

 moż

-1.34

 including

-1.27

by

-1.25

 where

-1.23

 کنید

-1.23

Not

-1.21

 between

-1.20

 pekan

-1.20

POSITIVE LOGITS

商品説明

1.40

 alteration

1.38

 regler

1.34

щаться

1.33

久しぶりの

1.31

{?>

1.30

 EVER

1.29

戬

1.28

 инструкция

1.28

 outset

1.27

Activations Density 0.070%