INDEX

Explanations

role of

The neuron is triggered by occurrences of “role of…” phrases that introduce what someone or something’s function or part is.

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

entry

0.66

ental

0.66

vans

0.66

janja

0.63

vers

0.62

VAN

0.62

rach

0.61

apes

0.61

ENTRY

0.61

lans

0.61

POSITIVE LOGITS

 निभाने

1.12

 Roles

0.97

 rôle

0.95

 desempeñ

0.94

役割

0.91

 desempe

0.90

 roles

0.90

扮演

0.84

 Role

0.84

ROLE

0.83

Activations Density 0.158%