INDEX

Explanations

fishing and hunting

The neuron is picking up on descriptive, product-feature language – the phrases and words that detail gear capabilities (lengths, stealth, abilities, etc.) in a promotional context.

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

Ingestion

-0.94

 nesting

-0.83

Runs

-0.83

öhnen

-0.79

inten

-0.77

 Runs

-0.75

 activités

-0.73

bite

-0.72

 spawning

-0.72

job

-0.72

POSITIVE LOGITS

 catching

1.34

 catch

1.33

 catches

1.19

 Targeting

1.15

Catching

1.13

 targeting

1.09

 CATCH

1.00

 caught

0.98

catch

0.98

 Catch

0.98

Activations Density 0.018%