INDEX

Explanations

where things stand

This neuron fires strongly on locative‐status expressions—phrases like “where X stands/stood” or “where X is on” something that describe someone’s or something’s position or status.

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

 will

-1.38

 bukanlah

-1.37

 horaires

-1.36

 réunions

-1.34

craper

-1.33

 chaussures

-1.31

 justru

-1.31

 économies

-1.30

 Nichts

-1.30

 понравился

-1.28

POSITIVE LOGITS

on

2.19

我又

1.41

ﾋ

1.39

now

1.30

 this

1.28

 investiga

1.25

你了

1.23

 tecnológicos

1.21

𝘵

1.20

Activations Density 0.014%