INDEX

Explanations

based + noun

The neuron activates on occurrences of the word “based” in compound adjectives (e.g. reaction-based, algorithm-based, concept-based).

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

thodox

-1.76

aneity

-1.75

ogues

-1.70

ités

-1.59

thodoxy

-1.53

íng

-1.49

ativement

-1.48

nesses

-1.46

???:

-1.45

mouthed

-1.44

POSITIVE LOGITS

ならでは

1.55

landır

1.53

べし

1.52

德國

1.52

 alkoholu

1.52

íguez

1.44

améli

1.43

 状態

1.38

 ANCHE

1.37

の商品

1.37

Activations Density 0.109%