INDEX

Explanations

legal action and lawsuits

This neuron detects language signaling the initiation or pursuit of a legal claim (e.g. “action,” “sue,” “recover,” “complaint,” “suing”).

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

bungen

-1.04

 Eingang

-0.96

自制

-0.95

ésor

-0.95

klju

-0.95

 zůst

-0.93

།

-0.92

Amigo

-0.92

титься

-0.91

 během

-0.91

POSITIVE LOGITS

 against

1.59

for

1.23

to

1.19

 damages

1.13

 seeking

1.09

pravy

1.07

וּ

1.05

ogues

1.04

 gerai

1.00

 dėl

0.99

Activations Density 0.080%