INDEX

Explanations

pre-trained models and embeddings

The neuron is triggered by jargon around pretrained deep learning models and transfer‐learning workflows.

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

 erlaubt

-1.48

for

-1.46

 también

-1.45

ைக்

-1.42

 szybko

-1.40

ﻂ

-1.38

ppsala

-1.36

 Когда

-1.35

 When

-1.34

"",

-1.34

POSITIVE LOGITS

 spektak

1.32

getting

1.30

铋

1.22

</h2>

1.21

 thiệu

1.20

taxia

1.20

ologically

1.15

 Junho

1.13

 gier

1.11

 które

1.10

Activations Density 0.020%