INDEX

Explanations

CPU specifications

The neuron activates on mentions of concrete objects or tools (e.g. pens, cookware, utensils, software tools) used in instructions.

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

imbulkan

0.76

 prestations

0.75

 Consultancy

0.72

ellingen

0.69

énération

0.69

 presentato

0.68

ettin

0.68

addam

0.67

urcharge

0.66

 Leistungen

0.66

POSITIVE LOGITS

 preferably

1.07

または

1.06

 или

1.01

 किंवा

0.94

 hoặc

0.92

หรือ

0.92

preferably

0.91

 alebo

0.91

 અથવા

0.90

もしくは

0.90

Activations Density 0.950%