INDEX

Explanations

comparison and evaluation

The neuron mainly activates on third-person singular present-tense verbs (e.g. “has,” “provides”).

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

 demás

-1.09

sonance

-1.05

análisis

-1.04

ITTLE

-1.04

 fieles

-1.03

vedì

-1.01

 salmón

-1.00

URIDAD

-1.00

ormick

-0.99

kington

-0.99

POSITIVE LOGITS

 more

3.06

 better

3.05

の方が

2.75

 superior

2.75

のほうが

2.41

 более

2.39

 greater

2.38

 easier

2.30

 stronger

2.27

 higher

2.19

Activations Density 0.114%