INDEX

Explanations

i would love to

The neuron strongly fires on first‐person statements expressing intent or willingness to do something again (e.g. “I’d use them again,” “I’d also like…”).

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

 haberse

0.79

Have

0.77

have

0.76

 hätten

0.73

mıştı

0.73

 telah

0.72

 hätte

0.70

Try

0.70

厸

0.70

 হয়েছে

0.69

POSITIVE LOGITS

 partic

0.69

 particularly

0.64

 cave

0.62

டிக்க

0.61

 subspecies

0.60

 arena

0.59

 aircraft

0.58

 airplane

0.58

 характеризу

0.58

 cumulative

0.58

Activations Density 0.075%