INDEX

Explanations

for surface adsorption

This neuron activates on the connecting words (prepositions and auxiliaries like “for,” “of,” “to,” “as,” “in”) that commonly structure titles and abstracts of scientific research papers.

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

 Siapa

-1.33

 renforcer

-1.31

 السياس

-1.25

ucap

-1.23

 Manfaat

-1.16

 glisser

-1.16

ureusement

-1.16

 большинстве

-1.13

irlo

-1.13

 използ

-1.13

POSITIVE LOGITS

for

1.31

 both

1.24

on

1.20

 purposes

1.20

 jako

1.16

1.10

始める

1.09

1.07

関する

1.05

 sowohl

1.05

Activations Density 0.248%