INDEX

Explanations

according to

The neuron strongly activates on the word “according” (as in the phrase “according to”).

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

などが

-2.52

齟

-2.19

しかし

-2.16

 ornate

-2.14

-2.13

在意

-2.11

だけど

-1.99

and

-1.98

では

-1.98

-1.95

POSITIVE LOGITS

蕯

2.66

kemeja

2.30

 lontano

2.23

鄚

2.19

 semasa

2.19

࣭

2.16



2.13

 anden

2.09

 sebag

2.05

 ditem

2.05

Activations Density 0.018%