INDEX

Explanations

troubleshooting or diagnose

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 excited

-0.12

 excitement

-0.11

 exhilar

-0.11

 admiration

-0.11

 exciting

-0.11

 congrat

-0.11

 celebrated

-0.10

 celebr

-0.10

 approving

-0.10

 enjoying

-0.10

POSITIVE LOGITS

 diagn

0.34

 diagnose

0.34

 diagnosis

0.33

 troubleshooting

0.30

 Diagnosis

0.29

 diagnosed

0.28

 diagnoses

0.28

 diagnostic

0.28

 solving

0.25

 troub

0.25

Activations Density 0.457%