INDEX

Explanations

not found or doesn't exist

This neuron detects error messages reporting missing or unavailable items, particularly phrases like “not found” or “does not exist.”

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

 flera

-1.24

UnknownFields

-1.12

阄

-1.10

 télécharge

-1.07

 perpé

-1.07

 largely

-1.04

 réuss

-1.02

 apporter

-1.02

 empê

-1.02

mainly

-1.00

POSITIVE LOGITS

or

1.52

in

1.49

any

1.43

all

1.34

 adequate

1.20

 specified

1.10

 anymore

1.09

at

1.06

nor

1.02

 whereabouts

1.00

Activations Density 0.037%