INDEX

Explanations

This neuron detects county names (capitalized place names) inside HTML <option> entries.

city names

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

ఝ

-1.15

goed

-1.11

 camiones

-1.08

 jejich

-1.07

裢

-1.05

ってました

-1.05

 ſet

-1.05

icky

-1.03

Ook

-1.02

WIL

-1.01

POSITIVE LOGITS

 with

1.44

at

1.33

on

1.27

 through

1.16

 after

1.15

 Although

1.13

for

1.10

 although

1.09

 traditionally

1.06

 during

1.03

Activations Density 0.000%