INDEX

Explanations

trust, rely, count, forward

The neuron detects expressions of trust or reliance, such as “trust,” “rely on,” and “count on.”

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

沨

-1.38

its

-1.35



-1.34

ulite

-1.30

 UNDER

-1.27

 aktual

-1.23

 Under

-1.23

pshot

-1.20

tasse

-1.20

 ‘‘

-1.19

POSITIVE LOGITS

 that

1.56

ֶה

1.43

 düğün

1.34

ようやく

1.27

Degrees

1.23

公式サイト

1.23

そろそろ

1.23

at

1.22

いよいよ

1.21

 Distributors

1.21

Activations Density 0.011%