INDEX

Explanations

purpose or direction

The neuron activates on words that denote coming together or cooperative action (e.g. “together,” “joined,” “in concert,” “around the clock”).

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

 tiroirs

-1.52

 bezpośrednio

-1.50

 drenaje

-1.49

 umożli

-1.47

私たちは

-1.45

 kérd

-1.44

our

-1.44

 zwięks

-1.43

 acogedor

-1.43

 sellado

-1.41

POSITIVE LOGITS

to

7.34

เพื่อ

2.13

 чтобы

1.83

 để

1.71

for

1.66

 لإ

1.51

collier

1.49

 щоб

1.42

bigli

1.38

 miesto

1.36

Activations Density 0.227%